Skip to main content

Full text of "Report of the Proceedings of the Conference on "Nine-Dot Braille""

See other formats


Proceedings of the Conference on 


June 19, 1964 


15West 16th street, NewYork.N.Y. 10011 










Dr. S. C. Ashcroft, Chairman 
Department of Special Education 
George Peabody College of Teachers 
Nashville, -'■'ennessee 

Mr. M. Robert Burnett 

Executive Director 

American Founiation for th*^ Blind 

New York, N.Y. 

Dr. Thomas A. Benham 
Professor of Physics 
Haverford Coll'-'ge 
Editor, Sci'^nce for the Blind 
Haverford, Pennsylvania 

F;r . John Covici 

Assistant in Developing "Nine-Dot- 

Braille"-Undergraduate Student 

■\'ay n e University 

Detroit, Michigan 

Dr. Emerson Foulke 
Deportment of Psychology and 
Social Anthropology 
University of Louisville 
Louisville, Kentucky 

Mr. Kenneth R. Ingham 

Research Associate 

Massachusetts Institute of Technology 

Cambridge, Massachusetts 

Mr. Howard M. Liechty 
Managinp- Editor of the "New 
Outlook for the Blind" and the 
"Matilda Zi egler Magazin<*" 
Member of f^e Board of Trustees 
American Foundation for the Blind 
New York, N.Y. 

Dr. Abraham Nemeth 

Assistant Professor of Mathenat-'cs 

Unive^^sity of Detroit 

Detroit, Michigan 

Dr. Carson Y. Nolan, Director 
De-oartment of Educational Rese-.rch 
American Printing House for the Blind 
Louisville, Kentucky 

Mr. Harold G. Rob'^rts, Director 
Program Planning Denartrnent 
American Found iti on for the Blind 
New Yo^k, New York 

Mr. Carl T. Rodgers 
Pro-ram Srecial ist in Braille and 
Other Tactual Education M.^.tcrials 
American Foundation for th? Blind 
New York, N.Y. 

Dr. Robert A. -cott 

Pro.iect Director 

Socia"! Science aid '/fork for the Blind 

Russell Sage Foundation and The New 

York Association for the Blind 

New York 

Mr. Robert Strom 
Inventor and Irotagonist of 
"Nine -Dot -Braille "-Under graduate 
Student-Harvard College 
Cambridge, Massachusetts 


Most people are aware, at least in a general way, that there 
exists a system of raised-dot writing which is read by the fingers 
of persons who do not see well enough to perceive the shapes of the 
letters of the Roman alphabet, and that this system was devised 
more than a century ago by Louis Braille, whose name it bears. Too 
often, the emotionality evoked by blindness obscures the most 
fascinating characteristic of Braille's invention. As the following 
illustration shows, it was designed and developed with mathematical 
precision and symmetry, for the inventor was an accomplished blind 
mathematician and teacher. 

The June 196^ Conference on "Nine-Dot Braille" was a logical 
continuation of the century-long efforts of Braille and his succes- 
sors to put the printed word, and all other written terminologies, 
within the reach and comprehension of those members of our literate 
society who are blind. 

It is also a logical parallel that a number of those who 
participated at the Conference should be accomplished blind mathema- 
ticians and teachers in various institutions of higher learning, for 
the braille system has always been to a large extent a function of 
education. It is probable that modern science and technology made 
.these participants more keenly aware of the magnitude of touch reading 
problems than were Braille and his immediate successors. 

The complexities of the task before the Conference are perhaps 
best illustrated by the second item on the agenda, entitled "Reada- 
bility of Nine-Dot-Cell Characters; Type Scale; Tactually Discriminable 
Space Difference Between Modifying Dots and Base Sign." 

It was not the purpose of the Conference to seek specific 
answers to any of the problems of tactual legibility, but rather to 
identify all the problem areas concerning the expansion of Braille's 
six dot base cell into a nine-dot base cell, and reduce these and 
other tactile problems into researchable areas. 

Mr. Strom's resume, which includes the reasons why he became 
interested in designing a "Nine-Dot Braille" code, and the agenda 
topic entitled "Need for 'Nine-Dot Braille'; Orthographic Efficiency 
and Space Saving," constitute an ample explanation as to why the 


American Foundation for the Blind considered the subject of 
expanded braille codes so important as to have called the Con- 
ference presently being recorded, as well as a previous one which 
was also held at the Foundation on July 6, I962. 

The Conference proceedings have been arranged in topical 
rather than chronological order to give maximum relevancy to 
each of the participant's comments and recommendations, and in 
keeping with the topics listed on the agenda which have not, there- 
fore, been repeated in this report. Nothing, however, has been 
changed, added, or deleted from the participant's statements. 

A Working Paper (see Appendix A) was prepared and sent to 
the participants in advance of the Conference. It was intended 
merely to stimulate discussion. Subsequent to the Conference, 
Dr. Nemeth offered to supplement his remarks on the determination 
of the intrinsic ambiguity of six-dot and "Nine-Dot Braille" 
(see Appendix C), 

The soundness of a philosophy of the presentation of 
written terminologies in tactile reading form, the logic and 
precision of the physical development of punctographic codes, the 
strict adherence to information theory in the assignment of mean- 
ings to the various characters of a code — these and other aspects 
of code development are highly important; but code optimization 
will not be feasible until the necessary data on the tactile human 
sense have been gathered and organized by research specialists. 

It was the hope of the Conference participants that 
extensive scientific investigations on touch reading would grow 
out of the Conference discussions. In the interval since the 
Conference, at least one major research study has been started. 
Dr. Emerson Foulke, of the University of Louisville, a participant 
in the Conference, has received a grant from the U.S. Office of 
Education for a project entitled, "The Development of an Expanded 
Punctographic Code for Tactual Communication." 

The study, according to Dr. Foulke, is not being undertaken 
with the thought of replacing the present braille code, but to 
determine the advantages and disadvantages of an expanded code. 
Many of the Conference participants will serve in advisory capaci- 
ties during the project. 


The American Foundation for the Blind, which provided some 
financing of the project, is pleased that one research project 
is already underway and hopes that still others mav be undertaken. 

The Foundation wishes to express deepest appreciation to all 
of the participants for the diligence and vigor with which they 
addressed themselves toward fulfilling the purpose of the June 19, 
196A- "Nine-Dot Braille" Conference, 

M. Robert Barnett 



Digitized by 

the Intel 
in 2013 










List of Participants 

Introduction ' Pages i-iii 

Opening Remarks (by Harold G. Roberts) Page 1 

Welcoming Remarks (by M. Robert Barnett 

Executive Director) Pap-es 1-2 

Resume on "Nine-Dot Braille" Code Develop- 
ment (by Robert Strom) Page 3 


Readability of Mr. Strom's First Nine-Dot Proposal. Page 9 

Readability of Nine-Dot Cell Characters, Type Scale; 

Tactually Discriminable Space Difference Between 

Modifying Dots and Base Sign. Page 11 

Need for "Nine-Dot Braille": Orthographic Efficiency 

and Space Saving. Page 3^ 

Questions on Braille Code Design: Philosophy of Pres- 
entation of Information to the Braille Reader. Page 50 

Participants' Comments on the Duplication of 

"Nine-Dot Braille". P^g® ^^ 

Consensus of the Participants' on Structuring and 

Organizing an Effective Research Project on Six and 

"Nine-Dot Braille"; remarks concerning Funding. Page 65 

Consensus of the Participants' on the question of 

Publicity. Pag® S8 


"WORKING PAPER" - "Nine-Dot Braille" Conference 

Submitted by: Carl T. Rodgers 

Specialist in Education 
(Braille & Tactual Aids) 


Report on "Nine-Dot Braille" Conference held at The American 
Foundation for the Blind, July 6, 1962 


"The Mathematics of Six-Dot Braille and Expanded Punctographic 
Codes for Touch Reading and Writing". 

Submitted by: Dr. Abraham Nemeth 

Assistant Professor of Mathematics 

Mr. Roberts: I am Hank Roberts. I am Director of the Foum^^tion ' s 
Program '^'lanning Der.-rtment and I am f^oinp to be here with you mort 
of the ^ay. I hpvp no technical knov/ledp-e of the sub.-ject th^t v;ill 
be discussed here today, but I am extremely interested and, more 
than that, I'm very s^rateful, as all of us he-° are, for qll of you 
coming here and r^t"''-"'-C"' ri;dting in this discussion that, I am suT-e, 
will prove to be very helrful to many, many i^eople. 

Mr. Parrett : Everv time v/e have any kind of a meeting in the 
foimdation, about all I do around here is come in and welcome grou-ps 
of people. I have such a stock speech that th'- staff has h^-ard i^- 
so much that T do sometimes try to change it, but it usually go-s 
along the line that it i" not up to me to w.=>lcome you -^ n the se^-se 
of helping the Foundation as such. It is, rather for us to thank 
you for using our facilities from time to time to try to advance 
knowledge about s^^rvices to blind persons or to advance or improve 
technical aid. 

This particular subject is one that rennires considerable 
genius, and I would say, brilliance. Evidently in this room there 
is probably almost all the knowledge there is in the United States 
of this -narticular subject. Apropos of braille, I remember a little 
verse that was once re-irinted in the rUTLOGK ard it w^nt something 
like this: 

Down in braille alley, down on my knees 
Tryinp; to find my stylus 
Lord help me, please. 

Can't you just get a vision of that little boy crawling around look- 
inp- for the stilus. 

I am also glad that thir plan, this study, has been going on 
for some little while now, thanks to Bob Strom's initiative'' , and the 
extreme interest that was immediately shown by Miss Gruber and Carl 
Rodgers and John Dupress, The reason I am mentioning those two 
names in particulc-r is that some of you may be aware that Miss Gruber 
and Mr, Dupress are no longer with the Foundation. 

Peyord this, the only thing I can assure you as a c-roup, is that 

the Foundation is prepnred to pick up whatever your recommendations 

are. I notice on the agenda later in the dav, the obvious questions 

of where do we go from here? 

I'm strictly an amateur, but have had enough coaching from the 
Research Staff to Vnow that if research is to be done ripht, it can 
be quite expensive and can take a lot of time and involve a lot of 
headaches and blood, sweat and tears. Insofar as T can help arrange 
the financing and the facilities for whatever the next step the group 


recoHuaends should be taken, you can certainly count on me and the 
Board of Trustees, 

Mr« Roberts; .......Carl has worked very hard in planning this 

meeting for today and has really done an excellent job. I am sure 
you would all agree that it is with a great deal of pleasure that 
I turn the meeting over to you Carl. 

Chairman Rodgers : Thank you very much for your very kind words, 
Mr, Roberts. 

Before I introduce Robert Strom, who is going to give us his 
resume on his "Nine-Dot Braille" efforts, I would like to say just 
one thing if I may, pertaining to myself. 

I have attempted to prepare a Working Paper on Research Areas. 
Know ye all, I am not a research man, and I do not labor under any 
hallucinations of being a research man or being acquainted with 
semantics, the jargon of research people, which simply means that 
as we go along with this discussion, if some other categories other 
than those which I have set down should be included or be taken in 
a different order, I certainly will be glad to do whatever seems 
most feasible to the group from the standpoint of a research discussion. 

I would like now to introduce Mr. Robert Strom, who will sum- 
marlBe his efforts on "Nine-Dot Braille". 


Mr. Strom : Thank you, Mr. Rodgers. 

I want to give you just a brief historical background as to how 
this system originally got started and what eventually brought me 
before the American Foundation for the Blind with a proposal of a 
"Nine-Dot Braille" system. 

It all started in June of 196l, when Mr. Donald Barr, who was 
then director of Columbia University's "Science Honors Program", 
which was a classroom and laboratory pro2:ram for pre-college students 
with unusual scientific aptitude, requested my assistance for a 
special task, 

I was a computer programmer, and I was requested to give pri- 
vate instruction to Mr. John Covici, who at that time was a student 
in the Mathematics and Computers Section of the Science Honors 
Program. I was supposed to ascertain all the problems involved in 
a blind person's learning, writing, and applying the techniques of 
computer programmers so that my instructions could serve as a sup- 
plement to the standard lecture and laboratory prop;ram used in teach- 
ing John's classmates in the Mathematics and Computers Course. This 
lecture sequence was, for the most part, rather visually oriented. 

For me and for the people at the Columbia Science Honors Program, 
this was an introduction to the problems and limitations of a blind 
person's learning a skill like computer programs. It was from this 
encounter with "trying to teach a program that involved writing of 
technical symbols that the idea of improving the braille system, 
which eventually led to the "Nine-Dot Braille" proposal, was conceived. 

Then in the course of instructing John, I familiarized myself with 
braille, with Standard English Braille, Grade Two and with the 195^ 
edition of Dr. Nemeth's Code for Braille Mathematics. I examined these 
codes carefully, trying to ascertain the logic behind the choices for 
the configurations. Many seemed at first sight strange. 

I had the conception that any system of communication, whether 
it was verbal or written, should contain a certain amount of redun- 
dancy of information; that is, whenever two signs of distinct sig- 
nificance occur, they would be less likely to be confused if there is 
more than just one perceptual difference between the signs like, for 
example, in spoken English, the words "bat" and "pat", where only the 
voicing of the consonants serves as a cue for distinction, are more 
easily confused than "bat" and "cut", where there is a difference not 
only in consonant quality but also in the vowel. 


Yet, in examining braille, there were many situations where the 
only other graphic separation between meanings was one of position 
with respect to blank cells such as the sign for "were" and the sign 
for the left parenthesis; that is dots 2, 3, 5, and 6, where the dis- 
tinction of meaning is based upon position of the sign for "was" and 
the sign for "by" where the distinction is made on the basis of wheth- 
er there is a space after the contraction. Or there were some char- 
acters where the only difference in perception between them was their 
vertical position within the cell. These seemed to be more percep- 
tually confusing than situations where there were more distinctions 
between the contractions. However, the need for redundancy in this 
perceptual legibility problem and the need for space saving and time- 
saving contractions which speeded up reading often ran in opposition 
to each other and had to be balanced off. 

When there are only 64 possible character configurations, there 
is a bad problem resulting from the fact that you do not have very 
many configurations to choose from. The result is that in literary 
braille, where we have a large amount of cueing from the context, 
what happens is that the orthographic redundancy is sacrificed for 
the sake of space saving contractions and, hence, we have a large 
number of situations where the same sign is used to mean different 
letters or characters, depending upon surrounding conditions. 

In mathematical braille there is less cueing from the context. 
We, therefore, had to make more different contractions, and we ran in- 
to problems where there are multiple cell configurations five or six 
cells long to represent one symbol in mathematics. Also, one of the 
patterns that ran through both kinds of braille codes was that of 
three indicator dots followed by a six-dot base cell basically a nine- 
dot configuration in two cells. 

So this was my first examination of the braille system that was 
used at the time I was teaching John. There were many problems with 
orthography as far as representing a lot of the information in the 
computer course and in the course in matrix alegbra, which is also 
being taught. These were mostly problems of orthographic inconvenience, 

I decided to experi ion of the braille system, 

that is, a way of expanc...^^ „ ..,.._ ., „ possible one-cell config- 
urations by a factor of eight. I felt that it was possible to both in- 
crease the number of possibilities for space saving contractions and 
also increase the number of different configurations and, hence, elimi- 
nate the overlap where we have different meanings assigned to these 
same configurations of dots. 


What I did at that time was the following: I retained the sixty- 
four basic single characters of Grade Two braille and I also retained 
most of the formerly developed concepts of how to make contractions: 
whole-word contractions, part-word contractions, single-letter whole- 
word contractions, initial-letter and final-letter contractions, short- 
form words, and then I added dots 7» 8 and 9 to the right of the base 
cell. I transformed the contractions which consisted of dots 4, 5» 
6, plus a ceil, to a nine-dot contraction containing the same base 
cell plus the indicators in position 7, ^ and 9 of the same cell. For 
example, "ation" which was dots 6, l-3-^-5i became dots 1-3-^-5-9. 

On the basis of this and also on the basis of an elimination of 
all the situations where the same sign stands for two different config- 
urations, I drafted a first tentative copy of both a literary and a 
mathematical braille system, more or less on my naive intuition at 
that time, and I then searched for organizational aid in supporting 
the construction of slates so that I could produce some "Nine-Dot 
Braille" in sufficient quantity for testing it. 

There were many nroblems in looking for help before I came to 
the American Foundation. Part of the problem was a fear of a new 
system of braille that would just add to the long history of changes 
that have been made in braille and part to the prevalent conception 
that psychological experiments had proven that six dots were the largest 
number of dots that could fit into a single cell. 

As to the second problem, I consulted Professor of Psychology 
Dr. Jerome Brunner of Harvard, and he explained that the experiment re- 
garding threshold of immediate perception and the maximum number of 
single perceptual elements that could be perceived in a single glance 
did not apply to the limitation of the size of the braille cell. It 
merely meant that when subjects were presented with patterns of dots 
on a tachistoscope in the case of visual images or with an instantane- 
ous glance and tactual images, there was a certain limit to the sub- 
■iectr. ability to identify them as being composed of a certain number of 
dots. But this did not mean that a person could still not identify 
patterns composed of more dots. He illustrated with the example of 
Chinese, where some of the basic signs consist of more than nineteen 
strokes; yet people can read Chinese very quickly even though they may 
not be able to tell at a glance how many strokes are in the symbol that 
they read. 

After some effort at locating help, I eventually came to the 
American Foundation and the very dedicated help of Mr. Rodgers, and he 
arranged a conference of specialists in July 1962 at which some of you 
who are here now attended. You can recall the recommendations that 
were made at that time, which included the following: 


1. "Nine-Dot Braille" has potential for increasing the 
scope and flexibility of tactual reading. It may be 
said that it could become a successful general-purpose 
reading system. We must find out whether it can, by 
experimenting net with the present "Nine-Dot Braille" 
that was offered before the last Conference, but, rather, 
with a group of systems which share the "Nine-Dot Braille" 
concept. In short, we should not stick to that one 
tentative proposal of assignment of meanings that was 
presented at that particular conference, 

2. "Nine-Dot Braille" has potential weaknesses which may 
aggravate the problems of tactual legibility. For 
example, bunched contractions should be avoided. How- 
ever, it was not known whether the appearance of illegi- 
bility of these contractions was the fault of the con- 
tractions themselves, i.e., the dot configuration; 
whether they were the fault of the configurations com- 
bined with their meanings, v/hether it was the fault of 
the unevenness of the slates that had been produced in a 
hurry for that conference; or whether it was the fault 

of the spacing value, which was .080" used in that version 
of braille. There were many different reasons, and it 
was not necessarily the "Nine-Dot Braille" concept which 
was at fault. We did not know which. 

3. "Nine-Dot Braille", whenever feasible, should continue 
the system and the pattern of Grade Two braille as much 
as possible so that people who are currently reading 
Grade Two braille do not have to make a violent tran- 
sition to "Nine-Dot Braille" if the system is ever re- 
leased. In particular, what was implied was that the 
"Nine-Dot Braille" cell should be considered as composed 
of tv;o parts: a 3 x 2 submatrix of the cell, which is the 
basic six-dot cell, plus three indicator dots which would 
serve as a modifier to the six-dot sign that is inside 
the nine-dot cell. It was also implied that the pattern 
of initial-letter and final-letter contractions should 

be retained. 

4. Indicator dots should be placed at the left and not at 
the right of the base-sign part of the cell, especially 
in mathematics where we had situations, when it was at 
the right, that a sub-2 looked like a 2-sub when read 
left to right. 


To experiment with the spacing values, the Foundation created 
three different slates which we have here. The spacing values of 
these slates are .O85", .090", and .095" between dots in a cell. 
Revisions to the "Nine-Dot Braille" system have been made since the 
time of the last conference. Sample material of literary amd math- 
ematical braille codes were made on each of the slates. We have 
those. And the results were shown to Mr. Rodgers prior to this con- 
ference, and he imrr.ediately pointed out, after careful examination 
of the "Nine-Dot Braille"key and of the samples of material, that we 
had a problem in legibility again which may be a result of the pro- 
blem in spacing. 

Basically, it is the following: what would seem to be a logical 
division of the "Nine-Dot Braille" cell into modifier and base sign 
based upon the dot numbers may not always be psychologically sep- 
arated into modifier and base sign according to the pattern that we 
have created. For example, remember that dots 7,o,9 are now on the 
left and dots 1,2,3,^,5,6 are now on the right. We had a contraction, 
dots l-2-3-if-7-8 (dots 7-8 "p"). Thpt would seem at first glance 
to be a good initial contraction based upon a "p" namely dot 7-8 p. 
However, when you actually read the contraction under your own finger, 
it turns out to look like an "er" sign with a dot trailing off at 
the right, and this would be a confusing perceptual cue because "er" 
does not have any phonetic resemblance to "p". 

Of course, one might argue that you could still use it to re- 
present, for example, a "per" sign. However, this would violate the 
pattern that we have set up, namely, that initial-letter contractions 
should always have the cue based upon the initial letter of the word, 
that is the "p". However, if we want to build a new contraction where 
this particular configuration is based upon the "er" sign, we would be 
violating the pattern that the indicator dot is on the left and not 
on the right. So we have a dilemma here. 

The question that was raised is basically whether there is a 
natural psychological breakdown of a given cell into the parts, a six- 
dot and a three-dot subcell. Or is it just that the particular appear- 
ance to Mr. Rodgers may not be the same as the particular appearance 
of the cell to another blind person reading the system? We have to 
determine this fact. 


The second problem is what to do about this. Should we relax 
the constraint about keeping within th« same pattern precisely of 
Grade Two braille or should we modify the spacing values within the 
cell so that the "Nine-Dot Braille" is no longer a square cell but 
instead a cell where there is more spacing between the three dots 
at the left than between the six dots at the right in order to sep- 
arate the characters more easily. 

The problem is what kind of alternative solution with regard 
to overcoming the problem does this pose and how does this affect 
the future of "Nine-Dot Braille"? Basically this, combined with the 
problems of legibility, are among the things that I would like re- 
commendations from this Conference today. 




Mr. Ingha m; Could I make a couple of points, having been at the last 
Conference? The first thing I would like to point out is that a 
resume of the suggestions of the last meeting as read here, I think, 
conveys a wrong impression. That is that these suggestions were by 
no means that firm. I think the only firm statements made were that 
"Nine-Dot Braille" could be read as it stood then, and that it should 
be researched as quickly as possible by a large number of people to 
see what could be done. I mean, we could not really tell at that 
time, in a one-day meeting, what these values and problems were; whe- 
ther, for example, dots 7,8,9, should be placed on the left or right. 
This was not at all firm. I think this is slightly different. I 
wonder if Tom Benham has similar feelings. 

Dr. Benham : As I remember it, we didn't come up with any specific 
recommendations about the aspect of it. It seemed to me as though 
we were talking about the possibility of getting the material in the 
hands of a lot of people so that it could get some widespread eval- 
uation instead of just by us. 

Mr. Ingham ; The only unified impression we all had is that, as it 
stood then, we could definitely read it. 

Chairman Rodgers ; Mr. Liechty, as long as we seem to be reminiscing, 
would you care to offer any comments? 

Mr. Liechty : It's my impression that there were no firm recommend- 
ations. I think I agree with these two gentlemen who have said that 
it appeared very legible. I was, I recall, amazed at how quickly you 
people, who read by touch, picked it up and followed Bob Strom's 
reading and prompted him at times when he hesitated a little bit. So 
it seemed quite legible as I observed the reading; also that it should 
definitely be tested by many more people. 

Chairman Rodgers ; Now I would like to give a little sequel to that, 
particularly with respect to the testing by many people. 

It was, as I recall, the consensus toward the end of the con- 
ference that "Nine-Dot Braille" should not constitute a new system, at 
least insofar as literary braille was concerned. The point was made, 
I believe by Mr. George Meyers, that during the last forty years or 
so we extended the situation to include Moon type and other forms of 
raised lines; that blind people have been subjected so far to many, 
many changes which, from a surely pragmatic point of view, would in- 
volve a lot of interruption in the progress that has already been made 
with respect to the development of touch reading and that, therefore, 


at least insofar as the literary braille is concerned, it should 
remain pretty much as it is now, with the possible exception of 
putting the modifying dots at the left and as part of the nine-dot 
cell. Thus the contraction for "mother" would not read "m" plus a 
middle dot but, rather, would read "m" preceded by a closeup dot 
which formerly, in a two-cell system, would have been dot 5 

Mr. Inp^ham; Just a couple of more things. Bob, did your material 
at the last conference have the modifying dots? For example, in 
the case of "mother", was dot 5, which is traditional, following 
the sign then? 

Mr. Strom ; It was at the right originally. 

Mr. Ingham ; I think we rebelled against the idea. 

Chairman Rodgers ; Yes, they did definitely. 

Mr. Ingham; As far as the random suggestion we made about the 
nine-dot system replacing the literary braille, you are right there. 
But as far as any other kind of braille, such as math braille, we 
felt that there were no holds barred and really we couldn't tell 
whether the dot should be on the right or the left. Or, indeed, 
whether the present way of writing was at all an efficient way in 
"Nine-Dot Braille". 

Chairman Rodgers; Right; and, indeed, that would follow along the 
traditional development of braille anyway, because some of the time 
in mathematical braille the modifying dots are at the left and some- 
times there is a modifying dot, or indicator, at the right of the 
base sign. 



criminable "sface difference between modify i ng dots and base sign . 

Cha irman Rodders ; The first thing I see on the agenda is "Readability 
of "Nine-Dot Cell Characters," and under that we have "Type ocale, 
Tactually Discriminable Space Difference of Modifying Dots, that is, 
between modifying dots and the base cell. 

Dr. Nemeth, would you care to start the discussion? 

Dr Ne meth ; For the sake of uniformity, I hope that we can get a 
standard way of talking about dimensions of braille, because trying 
to read different ideas that people have is very confusing. 

There are essentially three dimensions. There is inter-dot, 
inter-cell, inter-line. Inter-dot means the spacing between the 
center of one dot and the center of an adjacent dot in the same cell. 
By inter-cell, I mean the spacing between the dot in one cell, and 
the corresponding dot in the other cell. Grade Two, this is about 

250 of an inch. In various things that I have been reading, people 
have sometimes been talking something like I6O thousandths, and then 

1 understand the spacing between the far side of one cell and the near 
side of the next cell. This is very confusing. We should stick to 
one nomenclature. 

Inter-cell spacing is 250 thousandths of an inch standard, that's 
all. Inter-dot spacing is 90 thousandths ofan inch, that's all. 
Inter-line spacing is ^00 thousandths of an inch. 

Covici : That's corresponding dots. 

Dr. Nemeth : That's right. You should always use this nomenclature.. 

Chairman Rodgers: To satisfy Dr. Nemeth and anybody else who may 
disagree with that nomenclature in the Working Paper, I used both. 
- In some instances, it may be necessary to use both for clarification. 
I personally prefer the corresponding dot system — distance between 
corresponding dots of adjacent cells and adjacent lines. 

Dr. Ashcroft : It isn't quite true that it is from the far side of one 
to the near side of the other. It's from center to center m the case 
of dots. 

Dr. Nemeth: That's true. 


Chairman Rodders; Yes, it's from center to center in the case of 
the dot . 

Dr. Nemeth; There's a fudge factor there, I agree. 

Dr. Ashcroft ; There's a good reason for that, I think, because of 
the base diameter of the dot. 

Dr. Nemeth ; There is also an intra-dot dimension that I did not 
^n?!^°^n ! f"*! ^°lff dimensions of braille; intra-dot, inter-dot, 
inter-cell and inter-line. Intra-dot means the base diameter and 

tM« if . .*^^?^ ^°" ^^^ everything else can be inferred from 
this. There is no clarity gained by reverting to another system, because 
people are mentally oriented to one kind of set of dimensions, aid 
r?MnV^^>,^r^^ °*^^'' measurement from a standard set of dimensions, and 
I think that If you want to have universal understanding, you must stick 
to one set of specifications. That's the first thing. 

The second point that I wish to make is this. When you prepare 
experimental material, I think it would be lethal to try to put all 
i°: i^^V^'*^^^^^^ in at once. I notice that in the material of which there 
is only one copy, which you people studied at the last session, you 
have regular standard English braille where the inter-dot spacing is 
90 thousandths and then you have a nine-dot braille where the inter- 
dot spacing is 80 thousandths. I don't know whether there's also a 
discrepancy between inter-cell and inter-line spacing. 

Besides that, the standard braille has I89 contractions and short- 
form words whereas "Nine-Dot Braille" has lots of other contractions. 
So this makes the comparison very difficult. No one knows what they 
are really comparing. "^ 

Chairman Rodgers: May I say a word as to that. In the first place, 

the first conference was not structured as a well organized research 

fu°^®^u J\*^® second, there was no nine-dot writing equipment other 

I' m! ! *^^* ^^^ ^*^°°^ ""^^ successful in having constructed for 
him. That meant 

Mr. Covici ; That's not quite true. 

Dr. Nemeth; That's not true. I could have sent you a nine-dot slate 
if you wanted it. 

Mr. Covici ; (She New York Point slate is 3x3* 

•Editor's Comment: The New York Point System is two dots high. Theo- 
retically there is no limit to the possible horizontal extension of New 
York Point characters; however, in practice no character extended horizontal" 
beyond four dots. 


Dr. Nemeth : Yes 

Chairman Rodgers ; That would do it, but it wasn't available to us. 

Mr. Ingham : That's minor point anyway. 

Chairman Rodgers : Then you have an assignment of going around and 
seeing if you can collect one, 

Dr, Nemeth : My question is why not? 

Chairman Rodgers : Your question has been noted, sirl There was no 
slate available period; other than the one which Mr. Strom had, which 
would have meant that he would have had to write the whole thing, the 
sample material and everything else; he would have had to peck it 
out on the one slate, and in order to save time so that the conference 
could take place at the assigned date, the standard braille, the 
regular braille material was written on a braillewriter and the pur- 
pose of the other material was simi'ly to get a consensus as to whether 
it might be feasible to carry on from there with respect to research, 
and so on. We are not going to decide anything here at this meeting, 
gentlemen, as to which spacing variables and which set of comparisons 
is going to be the best. That comes at the end of the agenda, where 
you're going to have to think of setting up a well structured research 
project . 

That is the reason for that. We agree with everything that Dr, 
Nemeth has to say and we hope that, as the research project gets 
underway, all of these universal norms will be adhered to and that a 
nomenclature will definitely be agreed upon. As it is now, there is 
no specific nomenclature agreed upon because if you read some records 
of investigations — for example, those of the Commission on Uniform 
Type — they give you those that Dr. Nemeth has proposed. But if you 
read other experiments and other records, it is a little different. 

Dr. Nemeth : Let me make another point, the system of modifiers and 
full cell bodies — I don't know what else to call them — modifiers 
and cell bodies, is a very good system when it works for saving space. 
We have, from Mr. Rodgers' spacing values, the fact that a 40-cell 
line of standard braille will be reduced to approximately a 32 — or 
a 33 — cell line in "Nine-Dot Braille", Am I right about that, Carl? 

Chairman Rodgers : It varies. I would prefer to use the cell totals 
per writing area. I think it was reduced, depending on the size of 
the type, from 1,000 to 725. 


Mr. Covici ; Incidentally, that was the lowest. 
Chairman Rodders ; Yes. 725. 

Dr. Nemeth ; Yes, but that is if you have high density braille, you 
know, with small inter-dot spacing, 

Mr. Covici : No. That was the worst. 

Dr. Nemeth ; I see. 

Chairman Rodgers ; That was .090" or to use your terminology of 
corresponding dots, .3^0". 

Dr. Nemeth ; I see. Anyway, the point is this, there Is approx- 
imately a 20-to-25 per cent loss of numbers of cell per line or per 
page or per area, or what have you, in that neighborhood, in that 
order of magnitude. Let me confine myself to a set of contractions 
in the order of magnitude of I89, like in Standard English Braille. 
Let's keep that variable constant. Then this means that your nine- 
dot system will have to be overall — ■ not in the specific case, you 
understand, but overall -- approximately 33 per cent more efficient 
just to keep the space you already have without any gain on it. This 
means on the average that you will have to have approximately eight 
to nine dot-cell combinations per line just to break even with 
ordinary braille, because if you were to write ordinary text or any- 
thing like Bob had on the previous one, such as circuit statements or 
ordinary text where the contractions come at random, you don't get 
anywhere near the number of eight to nine cell-forms per line. You 
see, if you have 32 cells and you want to get up to ^0, you need those. 

If you have this system, it's very fine to have these long lists 
of contractions, but unless you get eight per line, you're not going 
to break even in space with your old-fashioned six-cell braille. This 
is the most important point to consider, I believe, even superseding 
space values and what have you, because if you're not going to break 
even with ordinary braille as far as spacing is concerned, you're 
not going to have any advantage. I wish you people would kick this 
around a little bit. 

Mr. Ingham : I would like to, with respect to English braille. 

Dr. Nemeth ; All right, I would like to comment about mathematical 
braille. If you use the modifier-cell-body combination, you have at 
your disposal eight modifiers including the "empty" modifier which is 
no modifier at all. As you know, Ken, from being on the math com- 
mittee with me, we have what, some 30-odd braille indicators? 


Mr, Ingham ; At least. 

Dr. Nemeth ; Some 30-odd indicators and we need them all. And the 
nine-dot cell is not going to help you. It is going to help you im 
this respect: First of all, one of the things we have in our math 
code is the possibility of performing five alphabets and within each 
alphabet we have the possibility of designing ordinary type, script 
type, bold face type, or italic type, which means for each letter 
you have 20 possibilities which is far outside the limits of your 
"Nine-Dot Braille" modifier product combination, you see. So the 
trouble is they are just as bad" in mathematical braille as they are 
in regular braille. 

Mr. Ingham ; I Just want to point out that there is an over-simplified 
way of looking at it. There are certain advantages which the "Nine- 
Dot Braille" might have in mathematics. 

Chairman Rodgers ; Do you want to get those on the record, please? 

Mr. Ingham ; I can't think of them all offhand, 

Dr, Nemeth; You would have to have eight modifiers per line to make 
it advantageous, remember, 

Mr. Strom ; We have at least that. 

Dr. Nemeth ; In mathematics? No, you don't. 

Chairman Rodgers ; Would you care to delineate what those are? 
Remember that we're relating this right now to spacing of type scales; 
in other words, physical compactness and so on. Are you ready, Robert? 

Mr, Strom ; One point I wanted to make with respect to mathematics. 

Chairman Rodgers ; Because if you're not, I think somebody called me. 
Yes, Dr. Foulke. Go right ahead. 

Dr. Foulke ; I was a little bit surprised, I guess, to hear so much 
of the conversation related to literary braille. If I understand it 
correctly, the whole discussion of "Nine-Dot Braille" came up in 
connection with the need for a more elaborate symbology to express 
.ther languages, didn't it? Mathematics, computer? 

Chairman Rodgers ; Both, 

Dr. Foulke ; You see, this problem did not arise in any effort to 
solve any problem relating to literary braille. 


Chairman Rodgers ; Yes. Excuse me please. Let me tell you a little 
bit of the background. 

At the first conference, Dr. Foulke, it was emphasized that 
since the literary braille readers would probably rebel at any new 
system or at what would seem to them psychologically a more com- 
plicated system because of nine dots instead of six, it was decided 
that if there was going to be a "Nine-Dot Braille" system, the stage 
of development would begin at the technical levels; in other words, 
mathematics, music — the special code levels, shall we say. But 
at the same time the possibility of extending "Nine-Dot Braille" to 
the ordinary literary code was not excluded at all. 

That is point number one. Point number two. When Mr. Strom 
presented the revised "Nine-Dot Braille" code that he had prepared it 
seemed, at least to me, that because of there being no space differ- 
ence between modifying dots and the regular braille cell dots, what 
resulted was an entirely new configuration in a good many instances, 
and you will see that later on. Mr. Strom and I decided — right, 
Mr. Strom, you can correct me on this -- that the same blending of 
modifying dot with regular base characters would also result in 
mathematical braille and, therefore, while we do believe that the 
developmental stage, so far as the structuring of a system is concerned, 
would begin at the special code levels, for practical purposes we 
took his literary key as an example because the type scale problems 
are going to be the same both for mathematical and literary braille. 
So we just took that one as an example. Do I make it a little 
clearer now? 

Dr. Foulke ; A little clearer except that I am not really sure that 
the spacing problems are the same. For instance, it may be that space 
economy is a more important factor in literary braille than in other 
codes. Space economy is, of course, important, I'm sure, in all of 
the codes, but it may be relatively more important in literary braille. 
This means that in the case of literary braille certain other arrange- 
ments might have to be made. 

Mr. Ingham : Could I make a comment quite relevant to this? This is 
the fact that no mathematical text that I know of is strictly math- 
ematics. Unless you're prepared to switch back and forth between two 
different types of writing, you're just going to have to be able to do 
all sorts of literary material in the nine-dot system. So I think 
it's a foregone conclusion that you have to consider it as literary 
material as well as mathematics. 

Chairman Rodgers ; I don't believe you can separate the two. 


Dr. Nemeth ; Unless you have one of these keys where they have nothing 
but theorem and proof and those are the only English words in there! 

Dr, Nolan ; Let me make a more general statement, Mr. Rodgers. 

It seems to me that we are all here because there is a problem in 
communication; that the present system is not versatile enough to give 
us the degree of communication required, and that we want to consider 
a modification of the system, namely, going from six to nine dots 
as a possible way of increasing the versatility. The line of con- 
versation to this point has been to give reasons why this won't work. 
We are just on the threshold of the whole business. I think that if 
we're interested in exploring increasing versatility of tactual com- 
munication, it's not going to do us any good to say it won't work for 
this reason^ it won't work for that reason. We haven't even reached 
a point where we know whether the characters are legible. This is an 
elementary thing without even getting to the point where we use these 
symbols symbolically or we use the stimulus patterns symbolically. 
So to dwell on these very practical problems, which I admit are of 
great importance in the light of the very little information we have 
about the potential of "Nine-Dot Braille" - we haven't studied it all - 
I think is a little premature. 

This is the thing that I confess I often find irritating, because 
when people advance new ideas and want to explore, there are always 
those around who say, "Well it won't work." This has been the history 
of science. There are always those people who sit around and say "No, 
that won't work," but people go ahead and explore anyway in spite of 
these, and I think that our conference ought to pursue it in the 
latter light rather than in the former. 

Chairman Rodgers ; Agreed. 

Mr. Strom : I agree. 

Mr. Covici ; Yes. In the light of that, I think what should be done is 
to see what kind of system is readable and do a lot of research on all 
the variables and then see whether a book takes less space or whether 
it's easier to transcribe or whether it's easier to read, which is 
the main point. 

Chairman Rodgers ; If we have exhausted the topic of type scale as 
such, we can easily go on to another topic. Before we leave type 
scale, however, I would prefer to give each participant the opportunity 
to express his opinion with respect to the importance of spacing 
variables as regards "Nine-Dot Braille", because there is one very im- 
portant factor here; How large should the type be in order, on the one 


hand, to be meaningful to the reading finger, and on the other, how 
small may the type be in order at least to break out even with six- 
dot braille, be it literary or mathematical or both, from the stand- 
point of physical compactness? 

I'm going to start over here at my left and let each one express 
his opinion on the specific point of needs for "Nine-Dot Braille" 
development with respect to spacing variables. 


Mr. Strom ; Yes. The first point is in relation to what you said 
about getting "Nine-Dot Braille" to break even with respect to purely 
the matter of physical compactness. One thing that has to be agreed 
upon is that you cannot specify merely from knowledge of the formal 
measurements of your cell how "Nine-Dot Braille" is going to fare 
against Grade II braille with respect to comparison. It's always 
going to depend upon a large number of factors that are independent. 
I wish to mention that we had some discussion with respect to ex- 
perimenting with a new kind of breakdown of the cell, i.e., having 
two different kinds of inter-dot measurements, that is, not have a 
square cell; have one kind of measurement for separation between in- 
dicator dots and base-sign dots and another kind of separation be- 
tween the individual base-sign dots and consider these as two dif- 
ferent spacing variables within the cell. 

Chairman Rodgers ; For what purpose? 

Mr. Strom ; For the purpose of making it easier for the finger to 
discriminate between what is to be the modifier within the cell and 
what is to be the base cell. 

Chairman Rodgers ; The base character? 

Mr. Strom ; Right. And this thing came up as a result of a number of 
perceptual confusions which should be noticed on the key that every- 
body has in front of him, Mr, Rodgers, do you happen to have an 
example of an easily locatable instance of this? 

Chairman Rodgers ; If you will give me a little time I will look 
for it. 

Mr, Covici; I have one right here. The one for "her", 

Mr. Ingham ; Where is that? 


Mr. Covici ; It's on the second line, very near the bottom. It's 
the second column. 

Mr. Strom : Yes. 

Dr. Nemeth ; I have it, 

Mr. Strom ; Right. Does everybody have that? Okay. The idea is 
that since I had the notion that a lot of personal pronouns had 
been contracted into one cell, "her" should also be contracted into 
one cell. I chose to try out dots 7-9 followed by a letter "h". 
Now the trouble with this is that it is probably going to look to a 
lot of you like a "z" with a dot trailing it. 

Chairman Rodgers ; Yes. I have here examples of characters like 
regular braille characters with a spare. Look for the 7-8 dot "e". 

Dr. Nemeth ; Where would that be? Is this in some order? 

Chairman Rodgers ; Yes. I would suggest that you look for the equi- 
valent first, which is a little to the right, and that stands for 
the letters "ee". 

Mr. Inghajn ; Where is that? 

Dr. Nemeth ; It's "in column 2, the sixth entry down. 

Mr. Strom ; That's an "e" with the dots 7-8. 

Dr. Nemeth ; 7-8 followed by an "e" yes. 

Mr. Strom ; It probably looks like an "f" with something after it. 

Dr. Nemeth ; It does not. Let me make this comment, Carl, for once 
and for all. 

Chairman Rodgers : Wait a minute. Let's examine a few of these. Now 
we have a stenotypist here, by the way, and every word that you are 
saying is being recorded. I would suggest that we can't get into an 
argument of what looks like what to individual "A" and what looks 
like what to individual "B" except perhaps to say that "X" number of 
individuals will take the 7-8 "e" as an "f" with a spare and "X-Y" 
individuals will take it as something else, but certainly you have 
no differentiation there between the modifier and the real "e" that 
it is intended to be. 

Dr. Nemeth ; May I say something? 


Chairman Rodgers ; Yes, 

Dr. Nemeth : The comment is this: What Carl says is perfectly- 
correct provided no one has preliminary instruction. But if you 
know that the structure of the cell is modifier plus body, then the 
interpretation of the "ee" combination as an "f" with a trailing 
dot is not an error, it's a blunder. It just can't be interpreted 
that way. 

Mr. Ingham ; I could make some points here which might be enlightening. 
We have at the lab a machine which can braille with a density of 
8,000 dots per 10" squared and I was testing this question with this 
machine. I talked to Bob Strom a couple of weeks ago about it, and 
I suggested this very thing of making one column closer to the 
other. Then I went ahead and played around with this machine and 
thought a great deal about it and I decided that it would be ex- 
tremely harmful to do it. The reason simply is that I could not dis- 
criminate when there were several in a row between that separation and, 
say, the separation that would occur between two cells, I got highly 
confused, and I think we really should make the difference not between 
columns but between cells. Make your cells distinctly apart, and I 
don't think you would have any problem. 

Mr. Covici ; Two points. First, three contractions, as Abe said, at 
first glance could look like anything you want. But if you know that 
they're supposed to represent a certain symbol, whether the modifying 
dot is before or after, or there is supposed to be a blank or whether 
it's so and so when you read it, you just have a table type thing 
and naturally it comes out. When you're going along reading, you 
don't say "Well, this is a modifier and this is a body." Only when 
you first lean it, do you think this way, and it may be nice to think 
of it that way as a mnemonic aid. Even though the cell used in the 
first "Nine-Dot Braille" conference was not so well spaced as the 
present one, it was quite legible and people were amazed, and I was 
too, when I first saw it. I think that the original cell spacing 
might be included in one of the research projects among many other 
variables but I'm not sure of its value, I think this "x" decimal 
of an inch may be equal to zero with very little loss in legibility 
or in anything else. 

Chairman Rodgers ; I'm going to turn the floor back to you folks, but 
first I would like you to look at some things that I could not find 
before and I have them labeled here. Samples of "Nine-Dot Braille" 
characters that seem to equal regular braille characters but, of 
course, they mean something else. Look at the sign for "ist". 


Mr. Covici ; It's on the first page on the right hand column. 

Dr. Nemeth ; I have it. It's in the middle of the third column. It's 
the fifteenth entry down. 

Mr. Covici ; I found it. Yes, indeed. 

Mr. Ingham : It looks like an equal sign. 

Dr. Nemeth ; No. It's 7-8 followed by the "st" sign. Let me say 
this: Bob has been very consistent in making every one of his con- 
tractions three dots wide. Am I correct about this, Bob? 

Mr. Strom ; So far. 

Dr. Nemeth ; I mean I haven't been going down the list, but so far, 
they are. 

Mr. Strom ; The ones that are two dots wide are not contractions but, 
rather, regular letters of the alphabet. 

Dr. Nemeth ; That's right. So every contraction which is three dots 
wide is not subject to a misinterpretation of any kind, of the kind 
that you were talking about with the 7-B "m" versus the thing moved 
over to the right. 

Chairman Rodgers ; The "st" sign. 

Mr. Ingham ; What is the matter with the sign? 

Chairman Rodgers ; This sign looks pretty much like a "gh" sign 
followed by the letter "a" to me. 

Mr. Covici ; Only when you first glance at it, but when you look 
at it — 

Dr. Nemeth ; Modifier-body tells me 7-8 "st". 

Mr. Ingham ; Bob, what you should do is give me or 10,000 other 
people a couple of weeks to read this. We can't tell. This is the . 
worst kind of test. 

Chairman Rodgers ; Right. This is not a test; this is a discussion. 


Mr, Ingham ; But you can't gain any information. It's just faked. 

Chairman Rodgers : This is a discussion and then these things should 
go on the record, because while it's true that right now we have 
a very erudite group of people here who would be perfectly capable 
of indulging in beautiful mental gymnastics in order to read, the 
fact is that other people around the Foundation to whom I've shown 
these things seem to have gotten the same reaction. So I just want 
to bring that out and put it on the record, and with that I turn the 
floor back to you gentlemen, 

Mr. Ingham ; If you're asking us on our particular level if we can 
read this, the answer is yes. If you're asking us in comparison with 
these other people, I think you are losing. I think this stuff should 
be going to thousands of people in test form. 

Chairman Rodgers ; It's a matter for research. I think we are all 
in agreement with that; right? 

Mr. Covici ; Yes, This is an open question and subject to review. 

Dr. Nemeth ; May I make the following suggestion, that when you have 
this thing tested, you do not say, "Well, here it is. Can you read it?" 
It should be precisded rather by a preliminary informational session 
where the potential reader will be told that "When you have a nine- 
dot configuration, you should regard the first three dots as modifier" — 

Mr. Ingham ; No. 

Dr. Nemeth ; Just a minute. Yes. 

Mr , Ingham ; Go ahead, 

Dr, Nemeith ; Of course. You should regard the first three dots as 
modifier and the rest of the thing as the body of the cell. This may 
matke a difference. You have all seen the game where you ask people how 
to pronounce a word, you spell the word but you stop, you pause between 
the wrong group of letters and people pronounce it wrong-^ naturally. 
Or you have all seen this gestalt experiment where you have a picture 
of a vase. If you look at the vase intently it becomes the figure and 
the black silhouette becomes the ground, and so you see a vase. How- 
ever, if you are told to look at the blacked-in material as the figure, 
then what you see is two silhouette faces looking at each, other and 
what used to be the vase is nothing but the ground. What you see 
depends on the way you are told to look for something. This is essential, 


Mr, InRham : There is experimental work going on now at our lab on 
this type of thing. Pictures of different alphabets are being 
screened on the oscilloscope off the computer and i^eople are looking 
at them. All sorts of things are going on, like letters upside down, 
word combinations put in different order, and what has come out 
immediately is the fact that by looking at the word as a whole you 
can buy a lot, by looking at the cells in sections, you lose. If you 
start thinking in terms of modifiers with this conglomeration of 
dots after them, then you are going to lose. 

Dr. Nemeth : I agree. 

Mr. Ingham ; This is quite contrary to what you were saying. 

Dr. Nemeth : I agree, but what I'm saying, for example, I tried an 
experiment here at the Foundation and it started out as follows: 
Someone wrote full braille for the first two lines. This was a 
coherent text, you understand. You write regular braille for the 
first two lines; then the next two lines was written with leaving 
dot three out of every third cell and then the lines thereafter 
were just leaving dot three and six out of every single cell. Now, 
you know, iust by the context I was able to read this thing down 
and there was no dot three or dot six anywhere, but I was able to 
read several lines beyond where there was any semblance of braille, 
iust from context and interpolation. IVhat you say. Ken, is true. 
You should perceive something as a whole, but there comes a time 
when you have to make a deduction, and when you have to make a 
deduction then you have to know on what basis to make it. You have 
to cue yourself to keep going. 

Mr. Ingham : Okay. Maybe at first in a training session. 

Chairman Rodgers ; I would like to insert a comment of my own with 
res"pect to the deductions that we may expect of the ordinary reader 
to have to make. After all, Dr. Nemeth, you are not an ordinary 

Dr. Nemeth : This was only an experiment. I don't read like that 
either, believe me! 

Chairman Rodgers : Experimentally or in terms of speed, on the 
average, very few people read like Dr. Nemeth and other peor-le 
I've known. I could mention them at random. Mr. A, who could read 
as f=ist as he could talk. I know Miss Y who could also read as 


fast as she could talk, but that is not the average. So it boils dovm to 
the fact that you should not be expecting the reader to read through 
deductions rather than by cue, the perceptual cues actually on the page 
that he might utilize if he so wishes. 

Chairman Rodgers ; I've not heard from some of our participants who 
have done some very extensive research on tactual discrimination, 
on errors in oral reading of braille at the elementary grade levels, 
and so on, and I would like to get their reactions, particularly 
with respect to Dr. Nemeth's suggestions about how to orient the 
subjects to the type and purpose of the test they are to take. 

Dr. Nemeth ; They should learn to read in terms of one third plus 
two thirds instead of two thirds plus one third. 

Chairman Rodgers : That should be explained to them beforehand. 

Dr. Nemeth t To be told before, yes. 

Chairman Rodgers ; Before the test. I would like to get the reaction 
of Dr. Nolan and Dr. Ashcroft and those who have been sitting very 

Dr. Nolan ; I will cease to be quiet, on your request. First let me 
say I think we are about seven leagues in advance of the question 
of cells within dot spacing in the cell. This I think was the ori- 
ginal question that was raised and we have gone a little bit afield. 
Actually I think there are two problems here. One is having a stim- 
ulus configuration with such spacing that people are able to identify 
all its elements when they encounter it. This is one problem. 

Then there is another problem that has to do with symbolic use 
for these different stimulus patterns. With reference to the first, 
my guess is that using an increasing number of dots in the braille cell 
is going to cause us to make them further apart. This is particularly 
going to be the case with younger braille readers. 

With reference to the problem of indicating the presence or 
absence of a modifying dot, that is, whether a dot 7i 8» or 9 is pre- 
sent, there is already a cue. If a cell configuration representing 
common usage is present, the space between it and the previous cell is 
going to be greater and regular-within-cell spacing will occur when 
you have the modifier present. 


I wonder about using distance as a cue for the present modifier. 
We are working within a very small distance to start with and I 
wonder if people are going to be able to discriminate the small dif- 
ferences in distance that will be involved. 

Chai rman Rodgers : Do you want, Dr. Nolan, to go to the extent of saying 
that it is so highly probable that they would not, that it would not be 
worth experimenting for the purpose of determining whether "X" decimal 
spacing between modifier and base character would be discriminable . 

Dr Nolan : I think we are certainly going to have to study this. One 
of'the problems, if distance appears to be a feasible cue, is the mini- 
mum distance that is discriminable to a majority of braille readers. 
This must be defined so that all people can use the system. There are 
other cues that one might use. Dot height is a possibility. That 
would present some problems for the writing devices that you employ. 
You may run into some problems there. 

Dr. Nemeth : May I interrupt just for a point of information. How 
uniform is dot height from dot to dot in a production run book? Has 
anyone measured that? 

Dr. Nolan ; It's been measured. I can't give you the information. It's 
fairly uniform. 

Dr. Ashcroft ; It's on the order of ten thousandths of an inch, I would 
^^, We had great difficulty, when we did the spacing variable study 
control, with dot height at 15 thousandths of an inch and we never did 
succeed really. 

Dr. Benham ; By the time someone has sat on it, why, it's not very 
useful anyway. 

Chairman Rodgers ; Thank you. Dr. Nolan. I think all these comments 
are very valuable and we want to keep them on the record. They will 
be very helpful in building up a set of recommendations to guide the 
research. The reason I stumbled there is because I have no idea what 
you folks are going to decide later on as to how to initiate at least 
some semblance of a structure for a research project that will come 
later. We are a little bit ahead of the game insofar as anticipating 
how much the reader should be expected to read by the actual stimuli 
which are presented to him and their clarity or lack of it, and how 
much by mental deduction or mental gymnastics, whichever you want to 
call it. I wonder if Dr. Ashcroft would have a comment on that just 
to get it on the record, because invariably that is going to come up 


Dr. Ashcroft; Yes. I would like to make several comments. 

First of all. I'm going back to the studies of spacing that we 
did. As you know, we varied the inter-dot, inter-cell and inter-line 
spacing and came up with 27 sets of specifications. It seems to me 
that it did not matter what you did very much in spacing; people could 
read it and we really had relatively short reading experiences for 
them and they accommodated rather quickly. Although Dr. Nolan alluded 
two considerations for younger readers and perhaps the necessity of 
some larger inter-dot spacing, if I heard him correctly, one of the 
things that came out of our dot spacing study was that with younger 
readers there was a slight tendency for them to read better with 
smaller inter-dot spacing. 

Mr. Ingham ; Really? Smaller? 

Dr. Ashcroft ; Yes. 

Mr. Ingham; That's interesting. They could read better than adults? 
Is that what you are saying, with the smaller dots? 

Mr. Strom ; No. They could read better with larger dots. 

Dr. Ashcroft ; They could read better with smaller dots. 

Dr. Nemeth ; By smaller, you mean .085"? 

Dr. Ashcroft: Yes. I think it was .085". It might have been as 
small as .080". 

Chairman Rodgers ; I think the tentative conclusion is that the 
inter-dot spacing they read best was .090", as in the case of the adults 

Dr. Ashcroft: Yes. It was between-cell spacing. I beg your pardon. 
That's right. 

Chairman Rodgers ; In other words, the distance between the end of one 
cell and the beginning of the adjoining cell was ,123"? 

Dr. Ashcroft; Yes, but I tend to agree with Dr. Nolan that this is 
far down the line and it seems to me the first thing to do is to get 
some "Nine-Dot Braille" out in this form and have it read widely and 
extensively. People can accommodate and learn very rapidly. From 
the error studies that I have done, I think that there are many errors 
that we don't need to be concerned about. People bring meaning to the 
context and the context is a crucial factor in errors and controlled 
difficulty level. This is extremely important for children. We can 


cause any of us to make reading errors if we make the difficulty 
level high. If the difficulty level is appropriate to our reading 
ability we can keep errors to a minimum. I just don't think that 
this is anything to fear at this point. I would go ahead and get 
a lot of material and get it rather quickly. I think it will be 
read, and read with facility rather quickly. 

Mr. Covici ; Dr. Ashcroft, what was the degree of accommodation 
when the inter-dot spacing was reduced to .072"? 

Dr. Ashcroft ; There were some statistically significant differences, 
but they were small, and again I would emphasize that these people 
had a very brief time to accommodate, and I think that they would 
accommodate even better if given more time. 

Mr. Ingham ; That we could all read the first material at the last 
meeting is more than adequate proof of what you are saying. 

Mr. Liechty ; Yes. \ 

Dr. Ashcroft ; People can read anything if they just have the meaning 
to bring to it and have a little time to work at it. 

Chairman Rodgers ; All the figures on variable spacing which I pre- 
sented in the "Working Paper" were only for the purpose of pointing 
out the tremendous number of possibilities that there are for vary- 
ing these things, and I'm glad Dr. Ashcroft brought up the point 
about ability to accommodate to various spacing values, because if 
that is the case, with adequate experimentation it would seem to me 
that, should nine-dot-cell characters be proved desirable other than 
from the spacing values, "Nine-Dot Braille" might be very feasible 
from the standpoint of physical compactness. 

Mr. Strom ; Mr. Rodgers, I would like to put up a question before the 
Conference for tneir opinion. 

Chairmam Rodgers ; With respect to type scale? 

Mr. Strom ; Right. With respect to the priority in research of in- 
vestigations of spacing. The suggestion has been made that it may 
be the case that cells will have to be made not square but, instead, 
with two different kinds of spacing values within them in order to 
provide for necessary cueing relating to modifier versus dot. 


Then there is the other opinion — that it may not be necessary 
to change the spacing if the cell configurations are suitably en- 
dowed with meaning in such a way that the mind learns how quickly to 
adjust itself to identify the nine-dot characters as an entire sym- 
bol. Also, it may be that in faster reading, modifiers are not even 
thought of or that the concept of the modifier and the cell is not 
necessarily one that has to be adapted, but there may be some config- 
urations which could be pure nine-dot configurations without the con- 
cept of modifier. 

The question which hence arises is, should we first try to see 
how well we can get "Nine-Dot Braille" to be read within the scope 
of the spacing values? I would tend to think that, first, we should 
see what we can get within the scope of a square cell, adjusting the 
meanings approximately and seeing how well they can accommodate to 
the cueing the way the square cell is now. 

Chairman Rodgers ; Does anybody have any question about what Mr. 
Strom said? Is it clear to everybody? 

Mr. Ingham ; It's not clear to me in the sense that just before he 
made his point I thought the gentlemen, Dr. Ashcroft and the others, 
as well as myself, made the point that spacing between different 
columns was not only unnecessary but harmful and, therefore, that 
question I thought was left. I mean, in what I researched up at MIT, 
to use different column spacing is definitely harmful and I under- 
stood that the gentleman over here was saying that it didn't mean any- 
thing either: The small amount of spacing you could accomplish be- 
tween columns would not buy the reader anything. 

Chairman Rodgers ; Was it Dr. Nolan? 

Dr. Nolan ; I commented on that, I wondered whether the variation in 
spacing that was possible would be such that you could get discriminable 
differences in spacing. I also wondered if it is really necessary to 
actually involve different spacing between modifiers and the others 
because, in the presence of no modifiers, the space between cells would 
be greater than in the presence of modifiers. This would be your cue, 
reduced space to start with. 

Dr. Ashcroft; Vacant space. 



Dr. Nolan ; That's right, vacant space would be your cue. I'm 
getting back to the thing that Mr. Strom asked, is the initial thing 
to study discriminability within these patterns? How far must a 
dot be from one another in order for people to be able to recognize 
all of them present with whatever degree of accuracy is required? 
This, it would seem to me, is the initial step to be taken here. Yo 
can test this difference of the distances between cells and the use- 
fulness of this as an indication of the presence or absence of mod- 
ifiers and also the between-dot distances as it affects the ability 
of people to be able to recognize all the stimuli present within the 

Dr. Ashcroft ; Some of these considerations were studied in con- 
nection with the spacing study that was done. 

Dr. Nolan : There is some data on this already. 

Dr. Ashcroft ; We couldn't vary the inter-cell spacing very fat 
because dots ^-5-6 became part of a new cell with the next cell 1-2- 
3. \ 

Dr. Nemeth ; That's right. It has a coherence with what follows. 

Dr. Foulke; I would like to comment upon something that I think 
Mr. Strom either said or intimated. This is that so far we have con- 
sidered the possibility of "Nine-Dot Braille" as a kind of logical 
extension of the existing braille code, in which case we have been 
concerned with modifiers and how they might be used, and we have dis- 
cussed the possibility of a rule of consistency regarding such mod- 
ifiers which seems entirely reasonable. But there is, at least for 
the record, another possibility of considering the patterns that 
result from the use of a nine-dot cell as stimulus patterns in their 
own right, in which case we might be concerned with the individual's 
ability to make absolute identifications of these patterns without 
having to go through any kind of analytical process of interpreting 
a modifier plus a standard braille configuration. This could be done 
and this is another possibility which I think at least ought to re- 
ceive some discussion. 

Also I would like to make a couple of other comments on some of 
the other things that have been talked about. I agree emphatically 
with Dr. Nolan that much of what we are talking about is the kind of 
questions that can really only be answered by research. I think it 
is clear that the research can be done. I think it is obvious to 
everybody here that spacing variables — within-cell spacing, between- 


line spacing — are all quite important. But beyond listing at 
least some of the infinite variety of possibilities that are avail- 
able, the only resort is research to answer some of these questions. 
It also seems clear that if we are going to answer them, it's ob- 
viously going to require the training of some subjects in the use 
of codes that we might generate. We can't present them to people 
cold and expect them to give us any information about their utility. 

Chairman Rodgers ; Dr. Foulke, I think your statement of testing 
the reaction to a stimuli per se is an extremely important point with 
respect to the development of any orthographic system. I'm speaking 
not now as a research individual but as a mere ordinary average 
citizen, an ordinary reader. It may be true that we ought to read 
two-thirds by stimuli and one-third by deduction or vice-versa. 
But I would say that as far as I (and perhaps I speak for the average 
reader) am concerned, we want to have as much information available 
under our fingers as possible. I don't mean over-cueing or anything 
like that, but certainly the orthographic stimuli should be clear 
enough that we may avail ourselves of three-thirds of the whole set 
of stimuli if we desire, if the information is otherwise not clear 
to us. Also, I think we should do a little developing of codes in 
a more systematic way for the sake of the nine-dot cell concept per 
se and for the sake primarily, of course, of the reader. 

Mr. Ingham ; The point I just wanted to make, following Dr. Foulke's 
statement, was that I think the way to go about setting up those 
codes is to give a couple of people the assignment of writing codes, 
keep these questions in mind and submitting these codes, and one or 
two or perhaps all of them, should be chosen and distributed. I 
don't think these are jobs for a committee to even pass on in terms 
of right or wrong. 

Dr. Nemeth ; Every good research project starts with someone form- 
ulating a hypothesis. This subject is sufficiently complicated for 
us to formulate several hypotheses. I wonder if it would be useful 
for us to go around and ask people to formulate what their hypothesis 
is concerning several subjects in this area: First, the readability 
of "Nine-Dot Braille"; second, which space "Nine-Dot Braille" would 
occupy versus the space that six-dot braille now occupies; third, a 
hypothesis concerning a structure in which each "Nine-Dot Braille" 
character has its own meaning versus the structure in which each 
"Nine-Dot Braille" character is a two-component symbol of modifier 
followed by principal character, I think these are the principal 


I think we already have the consensus accepted that the precise 
decimal amount of spacing between dots and space, between cells, 
and between lines, is easily adjustable too, and that the only im- 
portant factor is dot spacing within cells, and I avoid the word 

Mr. Ingham ; I think we disagree there. It is the other way around. 

Dr. Nemeth : Isn't that what Dr. Nolan said? 

Mr. Covici ; No. The other way around. 

Dr. Nemeth ; The spacing between cells was not so important as the 
spacing within cells. 

Mr. Covici ; No. 

Dr. Nemeth ; Dr. Nolan, would you edify us? 

Dr. Nolan ; I don't think I said any of those things! I said that 
the problem was one of legibility of the stimulus pattern and that 
this depends on the space between dots in the cells. This is the 
principal question. Then you get into the question of what is the 
cue to show whether the meaning to be attached to the stimulus pat- 
tern is that of the old six-dot cell or the modified six-dot cell. 
In this, spacing had been suggested as the cue, one idea being that 
the dot 7, 8, and, 9 should be spaced further to the left than dots 
1, 2, and 3 are from ^, 5, and 6. And I also suggested that perhaps 
the absence of a stimulus, that is the distance between cells, could 
serve as effectively as a cue. 

Dr. Nemeth ; Yes. 

Dr. Nolan ; But the elementary problem is one of legibility of the 
stimulus pattern. 

Dr. Nemeth ; That's right. 

Dr. Nolan; Can you report what you have been exposed to? 

Mr. Ingham ; I think this is the point. 

Dr. Nolan ; Then you get into the problem of codes. What meaning 
do you attribute to these patterns? 

Dr. Nemeth ; I suggested four hypotheses. Is it useful for people 
to formulate sort of a first guess as to what — 


Mr, Ingham ; The only comment I had about your hypotheses was, Abe, 
at least two of them had already been answered and I thought the 
third one had been agreed upon* The first one, for example, read- 
ability, had been answered no. Everybody agreed at the last con- 
ference and people agree here that it can be read, 

Dr, Nemeth ; This is not a hypothesis. Hieroglyphics can be read 

Mr. Ingham ; So you make up a code and then you determine this, 

Dr, Nemeth ; No. In other words, when I say "readability", I have 
in mind what Dr. Nolan has in mind, tactual discriminability. By 
that I mean readability. Can you identify one nine-dot character from 
another nine-dot character? 

Mr. Strom ; What constitutes identifying a character? 

Dr, Nemeth ; By naming the dot numbers. 

Mr. Covici ; No, maybe not, 

Mr, Strom ; I have a feeling that that is not the best way — 

Mr, Covici ; I'm not sure. It's not clear to me that naming of dot 
numbers is too significant. It may be that at the beginning naming 
the dot numbers is helpful. But, I really don't know, 

Dr, Nemeth; May I clarify that? 

Chairman Rodgers ; Excuse me. I would just make a comment here and 
then I will turn it back to you and pretty soon we will have to sum- 
marize this whole thing because time is getting short. You see, this 
is the valuable purpose of a "committee". We are not even sure how 
to go about these things. So the most we can do is to identify some 
of the areas of research, 

Dr, Nemeth ; You see, here's what I want to get. I want to get into 
the heart of the matter, I want somebody to make a statement. I 
don't care whether it's true or false, but a statement as to what we 
think will be the result in certain areas of this thing. You can dis- 
cuss it forever, and you're not getting close to this idea. But 
before' you pan- do research, you have to staj:e a Jiypotjaesis,.^ Maybe it 
|Will be borne out, maybe it won't, that I d^n't kno w. That»e thai . 
purpose of research. But you also do research this way. Someone 


makes a statement at the beginning of a problem that as a result 
of this kind of an experiment something will be the case. Then you 
do the experiment. Then from the experiment you say, "Well, I was 
right", or "No, I was wrong". But you have to begin with a statement. 

Dr. Nolan ; I would like to comment. The problem with expressing 
hypotheses at this point is that your hypotheses must be phrased in 
terms of the actualities of the research situation. This involves 
resources, people, apparatus, and so on. How these things will 
come into being is unknown to us at present. 

I would like to suggest that if we were to only achieve an out- 
line of the questions to be studied and these could be broken down 
in a variety of ways and hypotheses developed relative to them, re- 
lative to testing, this might be a realistic goal. In addition, we 
do not know who is going to do this research, and for us to impose 
upon this unknown party or parties a format may be a little bit 
premature . 


The other thing noticed is that it is possible to increase, 
with a "Nine-Dot Braille" system, the contracting power of a 
literary braille and it is also potentially possible to eliminate 
some of the perceptual difficulties that result from the fact that 
within six-dot braille you have, on occasion, signs that are difficult 
to distinguish from one another which require more context cueing 
to discriminate between them, I'm thinking in particular of the 
lower signs; I'm thinking of the situations where the same sign 
stands for three or four different things depending upon positions. 
Sometimes where the meanings are particularly unrelated to each 
other, like when there is a contraction for "was" or "by". I'm 
also concerned with the fact that the so-called signs, like the 
italics sign, the composition signs combined with the punctuation 
symbols often provide a very large number of cells just for the 
process of indicating what kind of character is going on, and 
"Nine-Dot Braille" could put, for example, a double capital sign 
into one cell and double italics signs in one cell and reduce space 
along those lines. 

The original "Nine-Dot Braille" had a proposal to eliminate 
such extraneous things as the numbers, or such potentially possible 
extraneous things, as the number sign, which added one cell every 
time you wanted to use a number and also required re-duplication 
of the letters "a" to "j" as the numbers. With "Nine-Dot Braille" 
the contractions of six-dot braille can be put into a smaller space. 
Presumably there is more compactness of presentation on the braille 
page. Contractions which previously had half of an empty cell in 
them, such as dots 4, 5» 6 followed by a letter, could now be re- 
presented in one cell without this additional space, so we have in- 
creased the number of contractions; we have more compact represent- 
ation; we have reduced the number of ambiguities of other graphic 
representation, and we have the facility for making braille capable 
of expanding to the needs of the new symbols that have to be created 
in special fields - mathematics, music, and dictionary writing. We 
have the composition signs made more compact. 

Chairman Rodgers ; I think you made mention of all those already 

and we have enough material, Mr. Covici, do you wish to add anything? 

Mr. Covici ; Yes, I wish to add the following. We have to get around 
the six-dot barrier, and there are a number of ways to do this. There 
is a need for some improvement over the old system and a nine-dot 
system is the thing which should be researched very thoroughly. Then, 
if this proves sufficiently worthwhile it should be suggested to 
the braille authority, and so forth, as a new standard. 


Chairman Rodgers : All right. 

Dr. Foulke ; First of all, Mr. Strom, I gather that in the case of 
a math symbology, we're not primarily concerned with space saving, 
are we? 

Mr. Strom : No. In any system we are concerned with ease of 

Mr. Ingham : Compactness. 

Mr. Strom: We're concerned with compactness insofar as it helps 
the blind person put more books on his shelf. We're concerned with 
compactness in that the finger has less space to traverse in reading 
material and, therefore, makes it possible to be done in less time. 
John has mentioned here that multiple-cell mathematical represent- 
ations are extremely hard to read. 

Dr. Foulke : This is what I'm trying to get at. The problem that 
arose, then, is that with the braille code it was either impossible, 
or else unacceptably clumsy, to express the mathematical language 
that you had to express in the braille code. 

Mr. Covici : That's right. 

Mr. Strom : That's right. 

Dr. Foulke ; Then this is a case that you are attempting to make, 
isn't it? 

Mr. Strom : In the case of mathematics, yes. 

Dr. Foulke : That in order to accomplish the job, in order to com- 
municate what you needed to communicate, there had to be a larger 
supply of symbols available. 

Mr. Strom : That's right. 

Mr. Covici : Yes. 

Dr. Foulke: I wonder if Dr. Nemeth would comment upon that? 

Dr. Nemeth: Yes, I would love to. I would like nothing better. 
First I wo uld like to make a few a priori judgements as to the space 
saving. I think that if the nine-dot system were used, even using 
the contractions that Mr. Strom has now, the amount of space occupied 
would be from 15 to 20 per cent greater than the amount now required 
for literary braille. 


Mr. Strom ; Okay. Next a priori judgment! 

Dr. Nemeth ; Let me say why. There are I89 contractions in braille. 
Short form words consitute, I believe, 76. 

Chairman Rodgers ; Yes. 

Dr. Nemeth ; Carl, do you know offhand how many initial-letter and 
final-letter contractions there are? 

Chairman Rodgers ; No, not offhand.* 

Mr. Ingham ; I8O something. 

Dr. Nemeth ; No, no. Initial-letter contractions and final-letter 
contractions only, something in the order of 86 out of 189, "ity", 
"ation", and so on. The alphabetic contractions would result in 
no space saving. The one-cell part-word signs would result in no 
space saving. The only characters which would save space would be 
the initial-letter contractions and the final-letter contractions. 

Mr. Strom ; Plus any new 

Dr. Nemeth ; That's right. Plus any new contractions which you care 
to add. I would judge that if you were to use no new contractions 
but just Grade II braille, and use the nine-dot cell to represent it, 
you would get something in the order of magnitude of about 25 per cent 
expansion. But because you added additional contractions in your 
code, the expansion will not be quite as great. It will be, however, 
about 15 to 20 per cent expansion in number of linear feet that your 
hand has to travel to get the information. That's what I mean. 

Mr. Strom ; Wait a minute. There was a 25 per cent expansion in the 
size of the "Nine-Dot Braille" cell that resulted from the mere extra 

Dr. Nemeth ; That's right. 

Mr. Strom ; However, you're not counting even the fact that these 
39 initial and final-letter contractions save one cell in addition to 
the composition. 

*In six-dot Grade II braille, initial and final-letter contractions are 
two-celled contractions. There are 33 intial-letter contractions and 
1^ final-letter contractions, a total of 4? two-celled contractions. 
There are 6? one-celled contractions and 76 short-form words, for a 
total of 189 contractions and short-form words (abbreviated words). 


Dr. Nemeth ; That's right. Let me say what I wanted to say before 
but which Carl didn't let me say, I was saying that on a nine-dot 
line you would have 32 cells approximately, maybe 33, which means 
you would have to have an average of seven or eight initial-letter 
or final-letter contractions per line just to break even. For 
example, take the first two lines of, say, the Preamble of the Con- 
stitution: "We, the people of the United States, in order to form 
a more perfect Union, establish justice," — there isn't a single 
initial or fina*l-letter contraction in that statement. 

Mr. Strom ; You know this? 

Dr. Nemeth ; I was looking through it as I was reciting it. "When 
in the course of human events, it becomes necessary for one people" — * 
there's the "one" to dissolve the political bonds which have con- 
nected them with another " — there's still just the one initial- 
letter contraction, "and to assume among the powers of the earth, 
the separate and equal station". There's the "ation". There's the 
second one, and so on.* You see, you have to have eight per line 
to bring up — 

Mr. Ingham ; May I interrupt? 

Dr. Nemeth ; Let me finish my comment. I will talk long but then I 
will keep quiet longer! In other words, as I say, unless you get 
the eight initial or final-letter contractions per line, you can't 
hope even to break even. 

Mr. Strom ; Or new contractions. 

Dr. Nemeth ; All right, or new contractions. 

Let me tell you my experience with Grade Three, Grade Three 
braille, just a brief summary for those of you who do not know, is 
divided into seven lines of braille. The seventh line consists of 
the modifiers, but the other six lines of braille are the non-modifiers 
In Grade Three braille you do this; You put a ^, a 4-5, a 4-5-6, a 
5 and a 4-6; in other words, five possible modifiers in front of each 
of the six lines of braille. That constitutes 56 characters in the 
six lines of braille, and when you multiply this by four, this gets 
you 224 initial letter contractions in Grade Three. 

In addition to this, there are prefixes and suffixes, 

*The word "among" in braille contains the final-letter contractions 
for "ong", 


There are contractions for "pre" and "pro", and "ate" and "ite", 
and "bly", and things of this kind, which are part-word contract- 
ions, which are also of this variety, modifier plus principle 

I regard myself as a fairly competent braille reader and I 
read material in Grade Three, Grade Three is very highly contracted, , 
It has about five or six hundred contractions overall, and you have 
to read that system mostly by deduction. What you are essentially 
doing is a very rapid mental table look-up, because many of the con- 
tractions do not follow normal mnemonic devices. Only 26 of the 
characters are letters. The point is, contractions like the "and" 
contraction, a certain position turns it into "anyone", "anybody", 
"anything". And the "of" contractions of "off", "office", 
"official" and so on. In other words, what you are doing when you 
are reading it is engaging in a continuous mental table look-up op- 

I don't know what the optimum number of symbols is for ease of 
learning, and this is a matter for research, of course. But there 
comes a time when if you condense the information too much — to 
the point where you have to do this consecutive table look-up — 
you're going to lose efficiency of recognition. You are going to 
gain a lot of space. You would think that since Grade Three is so 
highly contracted it would affect the space saving of a proportionate 
amount over Grade Two braille, but there is a very rapid low of dim- 
inishing returns involved. For example, in Grade Two braille, I 
think empirical evidence leads to the fact that the Grade Two braille 
effects a saving of about 20 per cent in space over what you would 
have if you used simple Grade One braille, 

Mr. Ingham ; 25, 

Dr, Nemeth ; 26 per cent 

Mr. Ingham ; Yes. 

Dr. Nemeth ; In other words, you didn't save that much. Now, if you 
go to Grade Three by doubling the number of contractions, you add 
only a few more percentage points in the saving ratio. In other words, 
to effect an additional 20 per cent of saving you would have to add 
maybe 2,000 more contraction*. 

Mr. Ingham ; There is a good counter argument, but I'll save it 
until you're finished, 


Dr. Nemeth : Let me finish this. Anyway, this is the thing. There's 
a lot of diminishing returns involved here, so that what you gained 
in space in Grade Three, you lose in speed of reading. 

Now, there's also this: There's no question that the modern six- 
dot braille symbol is highly inefficient, and there is a need for 
another system. But I think that a revolution in the braille system 
that would effect a saving, over Grade Two of only 10 per cent is 
not worth the revolution. If you could do it by an order of magnitude 
of something like making the distance one half of what it is, that's 
worthwhile saving. But just to get a 10 per cent discount at the 
cost of such a revolution, I dnn't know. 

There's another thing. Let me come to mathematics now. In 
mathematics the problem of putting tables into braille would be almost 
hopeless in a nine-dot system, for the following reason. It's almost 
hopeless already when you have ^0 cells available, but when you have 
only 32 columns wide to accommodate some tables, you're in all kinds 
of difficulty. You see, you can't put two digits in a slnp^le cell, 
can you. Bob? 

Mr. Strom ; No. 

Dr. Nemeth ; Then you are constricting yourself to a 32-column or 
33-column page, whereas 40 or 42 cells is hardly adequate now. oe- 
cond of all, I suspect — and I'm making another guess — in mathem- 
atics at the level of algebra, in other words — 

Mr. Strom : A saving of space over what? 

Dr. Nemeth ; Line for line. In other words, every time you write a 
formula in six-dot braille and write on in "Nine-Dot Braille", if you 
can get one on a line, you can get the other one on the line most 
of the time. In other words, in mathematics there won't be an ampli- 
ficction. If you are dealing with a special field of mathematics 
like vector analysis, where letters are modified by being written in 
bold-faced type, then for that kind of mathematics there will be a 
saving in the nine-dot system over the six-dot system by a small 
percentage. There I would judge it might save 5 or 10 per cent, but 
nothing like 20 per cent or 25 per cent at the numbers that I heard 
mentioned. I would imagine this is the case. 

Mr. Covici : We measured with the -mathematics — 

Dr. Nemeth: What do you mean, you measured? How can you measure it? 


Mr. Covici t With the mathematics that you have there for the 1962 
Conference plus the space taken up by the Nemeth Code, I mean as 
far as linear feet of braille was concerned — 

Dr. Nemeth ; Wait a while. You see, you are being unfair. You 
must not measure a code book. You've got to take a mathematics — 

Mr. Ingham; He did it. He took text material and he found it to 
be a 40 per cent saving on your code, that is "Principia", 

Dr. Nemeth ; I don't know what kind of symbols are in "Principia". 

Mr. Ingheim ; It doesn't matter. He did it. 

Dr. Nemeth; All right, it doesn't matter. I would say that in 
certain branches of mathematics there would be a saving. In no 
case would the nine-dot system take more space than the six-dot 
system in mathematics as it will in English braille except for table 
writing. I don't know what you can do about that. But, as I say, 
the nine-dot system will be a saving over the six-dot system in math- 

I don't know how you solve this— the problem in mathematics is 
not just the proliferation of numbers of symbols. In mathematics, 
you have essentially a two-dimensional notation. Fractions are 
written above a line and below a line. Superscripts and subscripts 
are written obliquely above and obliquely below. Modifiers are 
written directly above and directly below the symbols with which they 
are concerned. In braille you have, for all practical purposes, only 
a one-dimensional method of representation, and the problem stems 
more from the need to do one-dimensional representation than it does 
from the lack of number of characters. Ken will recognize what I 
am saying, 

Mr. Ingham ; I have lots of comments on your arguments when you've 
finished. I'll bring them up then, 

Dr, Nemeth ; I know you have. Now, for example, my fraction in- 
dicators are necessary only because I'm going from two dimensions to 
one dimension. In superscripts and subscripts, indicators are ne- 
cessary only because I have to go from two dimensions to one dimen- 
sion. My alphabet modifiers and my type modifiers, on the other hand, 
are necessary because of symbol proliferation. 


Mr. Ingham : The problem is, Abe, you bring up so many points so 
that when I get back to reply, I can't remember them all! 

Dr. Nemeth : Anyway, the point that I want to make is that there 
will be a variation in saving mathematics by nine dots over six dots, 
depending on the system used, and the saving will be only in the 
m-.theraatical -portion of the material, which far outweighs the formulae 
portion . 

Dr. Foulke : Dr. Nemeth, let me ask you a very specific question. 

Dr. Nemeth : I h'lve , for all intents and purposes, finished. 

Dr. Foulke : As far as space saving is concerned, it seems to me that 
again this doesn't have to be cur primary requirement. If a nine-dot 
mathematics code will do a job for us that the six-dot code will not 
do, we can afford the paper. 

Dr. Nemeth : Y e s . 

Dr . Foulk e : But what I would like you tn answer specifically then 
is, will it do a job that the Nemeth Code will not do or will it do 
it significantly better? 

Dr. Nem eth : I don't think it will do a job that the Nemeth Code 
will not do, because it will also be based on a system of modifiers 
just like the Npmeth Code now is, 

Mr. Ingham : The real catch to his nuestion is the second part: i-/ill 
it do it? Will it do it better? 

D r. Nemeth : That I don't know, whether it will do it better or not. 

Dr. Foulk e: This, it seems to rae , is what we need to know. 

Dr . Nemet h: Better from what point of view? 

Kr . Covici : The reader's. I'm not concerned with the transcribers, 

Mr. In g ham : Can we hold the answer to the second -oart of the nue.-tion 
until I have my points made? 

Dr. N ^m^-th: 'fa^t a w^ile. Dr. Foulke h^s asked me a qu'='stion. 

Mr, Ingham : I know, but I'm just saying that it would be better if 
we c^ulr] hold it. 


Dr. Nemeth ; Go ahead, I would like to answer it later. 

Mr. Ingham; Yes. I want you to remember it because this is one 
of my points also. The first comment is, all of your arguments, 
Abe, are based upon the fact that the code in its best form is 
based upon the six-dot system. 

Dr. Nemeth ; Which code? 

Mr. Ingham; The nine-dot codes. This need not be the case. Just 
take, for example, your question of algebra and literary braille, 
things like that, where you said the savings would be equal or less 
than the six-dot braille savings over, say. Grade One, okay? 

Dr. Nemeth ; I don't know. I don't follow you; I'm sorry. 

Mr. Ingham ; You said that in the literary material you are going 
to lose with the nine-dot system. 

Dr. Nemeth ; Yes. 

Mr . Ingham ; In elementary material in mathematics you're going to 
maybe break even. 

Dr. Nemeth ; Yes, you're going to break even. 

Mr. Ingham; Now, I disagree with you on those points for the simple 
reason that there are situations which could be devised, with a 
proper choice of symbols, where we would save fantastically. Take, 
for example, your chart problem. Suppose your cells, as they often 
are, included numbers with signs in front of them, plus and minus, 
letters which were capitalized. 

Dr. Nemeth ; No. 

Mr. Ingham ; Just a minute. Let me finish. 

Dr. Nemeth ; You're not realistic at all. 

Mr. Ingham; I am so. You can have tables with lots of numbers and 
signs in front of the tables. Secondly, you can have tables, partic- 
ularly in programing where your letters are all capitals. All IBM 
manuals are typed with capital letters. When the transcriber gets 
it, she's going to type capital. Now, Mr. Strom can devise a system 
where his nine-dot cells include those symbols which in your system 
take up a total of four columns. 


Dr, Nemeth ; I agree with you for your capital signs, but I don't 
agree with you for your plus and minus. There's no way of incorp- 
orating a sign for a number with the number in the same cell, 

Mr. Ingham ; I don't agree with that. I don't see why Mr. Strom 
can't use a group of four dots, say, in the nine-dot cell and use 
an extra dot to indicate a minus sign in the absence of a dot plus 
a sign, for example. 

Chairman Rodgers ; I'm a little confused here and I want a little 
clarification for myself. Mr. Strom, have you provided in your 
revised code of mathematics for the incorporation of signs of op- 
eration with the number itself? 

Mr. Strom ; No, not in the current system. 

Chairman Rodgers : Do you think it would be feasible to combine a 
sign of operation with the number to which it refers?* 

Mr. Strom ; It could be, if I were convinced that it would be 
advantageous to do so. I don't know whether it is. 

Chairman Rodgers ; You don't know? 

Mr . Ingham ; This is the question, but it could be done. 

Dr. Nemeth ; Ken, do you realize that there are just now the two 
possibilities? You have to provide for the possibility of an un- 
signed number, a number with a plus sign, a number with a minus sign, 
etc . 

Mr. Ingham ; Go ahead. You still have nine dots to do that. 

Dr. Nemeth ; I know, but I have these numbers without — 

Mr. Ingham ; Let me just make one point. 

Dr. Nemeth ; Let me answer to this one point. Unsigned numbers, 
numbers with a plus sign, numbers with a minus sign, numbers which 
are preceded by plus or minus, and numbers minus or plus. All of 
these possibilities exist. 

*In Literary Braille there are special symbols that represent accented 
letters in foreign languages, thus combining letter and accent into 
one braille cell. See "English Braille American Edition 1959, Revised 
1962" - inkprint pages 22-2^ 


Mr. Covici ; Would the eight cells in the table make a difference? 
There are often a lot of tables where they put lines between the 
numbers, and they put crazy lines down the page, and they do all 
kinds of tricks. 

Dr. Nemeth : Let me ask you this: How would you do a simple table 
of logarithms: — no signs, just digit after digit? Ten columns, 
five digits per entry with a column of "n" on the left, you know, and 
across the top? What's involved? 

Mr . Ingham : So you use two pages. The thing is this: I agree 
with Dr. Foulke that in essence you may lose in certain situations 
such as you're pointing out in a table of just pure numbers, un- 
signed. Okay, fine. The question is, how would it do with the 
rest of the code? I think the point that Mr. Strom made with me over 
the phone is that he looked up in the American Physical Society's 
booklet, or whatever it was, a list of some 500-odd standard symbols 
which they recognized that all print books use, for example, in 
articles in their journals. With his system he could account for 
every one of them with a single cell, 

Mr. Covici : It may not always be possible. 

Mr. Ingham : This may not be always true, but the tendency is such 
that you can do this. 

Another thing as far as your code goes. There are situations in 
your exponent problem which is really one of the most important points 
in transcribing math, or other sciences, where these level indicators 
can be either included in the same cell or can be given a separate 
cell of their own, while you use two or three cells in between modi- 
fiers. I think all of this discussion is a little bit irrelevant due 
to the fact that you will have to admit that a code has to be de- 
vised, which can bring these things out much better than what we have 
before us here. We really can't say categorically how much it will 
save and how much it won't since we don't really know (l) how many 
of the signs can be read without ambiguity, and (2) whether Mr. Strom 
can devise a code with, say, the logic your code has which, will be 
superior to it. I don't think we can give an answer. This is use- 
less stuff; that you will end up, say, on an equal basis with 
alegbra and you will lose with literary braille. I think that there 
are possibilities where you can save quite a bit. 

Mr. Covici : And we did save ^0 per cent in the one case. 

Dr. Nemeth ; John, could I ask you a question. What kind of symbols 
were there? 


Mr. Covici ; You have the thing there to look at. 

Mr. Ingham ; There was everything in it; all sorts of weird 
assumptions, and so forth. 

Dr. Nemeth ; I don't see it. 

Chairman Rodgers ; Go toward the part of mathematics. 

Mr. Strom : I wish I had the key for the braille that was used 
two years ago, which is completely different. 

Chairman Rodgers ; Yes, it is, and this is why I didn't feel it was 
worth distributing. 

Mr. Strom : This is the six-dot braille? 

Dr. Nemeth : The six-dot braille. The small sheets are the six-dot 
braille. Where am I to look? 

Mr. Strom : I better point it out for you. 

Mr. Ingham ; I just have one more point to make in order to finish 
my 'statement. The point is that ^0 per cent was saved over your code 
with this kind of material. 

Mr. Covici ; In linear feet of braille. Thirty-six or thirty- 
seven per cent; I don't remember. 

Mr. Strom ; I can't recall the figure, 

Mr. Covici ; It was around there. 

Mr. Strom : I don't want to commit myself. 

Chairman Rodgers : If I may interject here, we could go back and 
forth with the ^0 per cent or no ^0 per cent for the rest of the 
afternoon. We don't have the tabulations; we don't have the necessary 
material step by step, and a conference like this would not be com- 
petent to judge whether a statement of that kind is correct at its 
face value and we would have to take it at its face value. 

Mr. Ingham : This is part of my statement. There's only one thing 
I would like to say. As far as Abe said, he brought up the case of 
Grade Three braille, and the saving it made, and so forth and so on. 


and I would suggest that much of the problem, as he pointed out 
with Grade Three braille, was that it involved the attempt on the 
blind reader's part to coalesce two or three cells or say, two 
cells, and I think you would win over a system using 500 or more 
contractions, as with Grade Three, if they were grouped in a one- 
cell configuration and you looked at it as a complete cell^ not as 
you insist on doing, Abe, as a cell with a modifier. 

Dr. Nemeth r Then you're doing this table look-up that I'm talking 

Mr. Ingham ; No, you're not doing any table look-up at all. You're 
designing a code which, as you learn it, will be just like the 
simple Grade One braille is now. 

Dr. Nemeth ; Only 512 instead of 26 letters. 

Mr. Ingham ; So what? 

Mr. Covici ; When you read it — 

Mr. Ingham ; There's nothing mnemonic in Grade Two braille as far 
as I can see. 

Dr. Nemeth ; There are a lot of things which are not mnemonic. I 
don't know where the breakdown begins. 

Mr. Ingham ; The point is there's no reason why dots 1-2-4 for 
example, should be "f" in mnemonics. There's no good reason for it. 
Once you have assigned a characteristic group of symbols to a word 
or a part word, there's your mnemonics. It's when you start going 
into more bits of information like two cells or more that you con- 
tinue to have to look up. 

Dr. Nemeth ; What is mnemonic is that "b" stands for "but", and 
not for "cow", that "f" stands for "from" and not some other letter. 
What you say is correct. These contractions in the code that Bob 
has are fairly mnemonic, whereas Grade Three wasn't, but this can 
be the case only if you stick to the modifier plus principal chara- 

Chairman Rodgers ; I think we will have to terminate this portion of 
the discussion. The main point which should have been established 
by now, which originated in Dr. Foulke's question was, is there a 
need for a "Nine-Dot Braille" system? Correct, Dr. Foulke? 


Dr. Foulke ; Yes. 

Chairman Rodgers ; I think enough has been said to enable the 
participants as a whole to judge for themselves and form a consensus 
as to whether there is or there isn't such a need. 

Dr. Nemeth spoke of the law of diminishing returns with respect 
to space-saving efforts. What I was concerned about, that Dr. 
Nemeth didn't mention, and again this is just for the record on be- 
half of the average reader, is that the more contractions you put 
into a system the more you increase the need for braille rules. 
Braille rules up to a point can be tolerated by the average person 
so long as they don't become too numerous. Let's suppose that we 
have a contraction for "ide" which does not exist in Grade Two braille, 
Does it or doesn't it? 

Dr. Neneth ; No, not in Grade Two, 

Chairman Rodgers : Okay, it does not. In Grade Three it does. 

Dr. Nemeth : In Grade Three there is a "de" and you can leave out 
the "i". 

Chairman Rodgers : Suppose there's a clearly defined contraction, 
the "ide" contraction for "chloride", "confide". That's fine, but 
when you come to a word like "confidence" 

Dr. Nemeth ; "Mideast"! 

Chairman Rodgers: "Mideast", I wonder what's going to happen to 
what Mr. Strom has described as the morphology of English words? 

Dr . Nolan ; It seems to me again we are way ahead of ourselves. 
We're talking about codes and the degree to which they will save space, 
I think the purpose of us being here is not to decide these issues 
and not to decide whether or not we will adopt "Nine-Dot Braille", 
but to decide how it might be studied. And all these things are a 
little premature, I believe. They are taking us far afield from 
our goal. They are pertinent problems but they are problems that 
are going to be an integral part of the study. The codes that we 
originate are certainly going to result from consideration of these 
problems that have arisen here. But this is for the future. We 
don't have the codes jelled now. In fact, there can be several 
codes. So wouldn't it be more to our purpose to get back to it? 




Dr. Nemeth ; First of all, let me put in a plug for modifiers on 
the left as being consistent with everything we know about information 
theory. Those of you who know my math code I now that one of its 
salient features is to give the reader information in advance of 
what's going on instead of afterward. So we have the fraction indi- 
cators to tell you that a fraction is about to happen, and we have a 
modifying indicator to tell you what the precise nature of the mod- 
ification might be until later. So we always have to have the 
information first. 

In fact, even in literary braille the people who devised the 
Spanish language have an advantage over us. They put the question 
mark in front of the sentence to let you know that it's going to be 
a question; this is a superior way of conveying information. If you 
put modifiers on the end, then you have no readiness for the mod- 
ification at all and you are likely to read it in its bald form 
before you realize that it should have been modified. That's the 
first thing. 


Chairman Rodgers ; Before we go into the second thing, may I 
interrupt you there for a second. You do have modifying dots at 
the right to close up a situation. 

Dr. Nemeth ; To show the termination. They are not modifiers. This 
is to put a limit on it. In other words, it tells you when a mod- 
ification is ended. 

Chairman Rodgers ; Right. 

Dr. Nemeth : But this is just so that you could now sort of clear 
your readiness for the next thing. 

Chairman Rodgers ; Good enough. 

Dr. Nemeth ; But the important thing is to know when modification 
begins. This is salient. This is paramount. It is important also 
to know when it ends, but more important to know when it begins^,.,. 
In addition. Bob was making a comment about 8-9 followed by a "p", 
I suspect that the psychological confusion — ■-■ 


Chairman Rodgers ; 7-8? 

Dr. Nemeth ; Yes, 7-8. I suspect that the confusion by which it 
was read as "er" followed by a trailing dot 4 was more the result 
of lack of familiarity with the system than anything else, because 
if one is used to a pattern of modifiers plus ^-5i modifier plus 
^-5, modifier plus ^-5, one will have a mental set to read the ^-5 
by itself and not to initially try to associate the other three 
dots with that. 

Mr. Ingham ; You see, this is one of the reasons I brought up the 
question of the last meeting, and the impression is again in those 
notes, I don't think that was a question that came up in the last 
meeting at all. In other words, we found that it was readable; 
there were no questions as to whether signs were confused with some- 
thing else. I think this is an intellectual question that has been 
raised rather than a natural practical situation which arose at the 
last meeting. 

Mr. Covici ; When the dots are on the right in some cases, it may 
not be a question of a mental set to read modifier plus body. The 
person may still read that more like a "p". Also I think that in 
questions of telling you what's going to happen, whether it's on 
the left or on the right, whether the modifier should come first to 
tell you what's happening, I'm not sure whether the time lag is really 
of too much significance; this would have to be researched. I could 
see that in a fraction, you would want to know when it actually 
begins. That's a whole level space, but whether you have it on the 
left or on the right in the cell, it is not a great distance in 
space. There's not a very great difference in time. I've seen 
material from the last conference. You read a sub-2, say. You read 
"a" and you read the sub-2 or most times sub-one. 

Dr. Nemeth ; That's a right modifier. 

Mr. Covici ; Yes, that's right, but the information doesn't come 
too late. 

Dr. Nemeth : No. 

Mr. Ingham ; I think Abe's point was well taken. He was pointing out 
that, for example, in the case of literary braille, dot 5 "m" tells 
you that the "m" is modified rather than "m" followed by a dot 5. 
Yo- don't really read that as a full thing or at least you haven't 
where it included two cells for the combination. It may be different 
with the one-cell case; I don't know. 


Mr. Covici ; Another point about Dr. ^emeth's system; in general 
systems of mathematics and of literary braille, there can be at 
least two different orientations, (a) the writer, as the person 
doing it; and (b) the reader. Dr. Nemeth's code is oriented de- 
finitely towards ease of writing by the transcriber. It's very 
easy to write and somewhat difficult to read. 

One symbol may be strung out over a large area. This impairs 
legibility of braille in mathematics notation. The "Nine-Dot 
Braille" notation would be more compact. Dr. Nemeth's system is 
perfectly computerized. In other words, it would take only a 
machine-type operation. But on the other hand, it makes it some- 
what harder to read, and I think as far as time is concerned, which 
is what we really want to save in the reading, the point of saving 
space is to decrease the time; I think it has to be definitely 
oriented toward the reader and I think it must be stressed. I 
don't think it can be overstated. 

Mr. Strom ; I think it is very important to mention that if you're 
going to work on designing a system that is in the interests of 
the reader rather than the transcriber (which has to be always kept 
in mind first) we should hesitate when we say we're developing a 
new code for mathematics to mention whether it's absolutely ne- 
cessary to have 10 to 15 indicators. Basically, as I understand 
it, the present mathematics code gives a good deal of information 
on how the imprint shape of the characters may have looked, which 
may or may not be important when you consider it in the light of 
merely being able to read the material more easily. It may be that 
cueing other than that based upon the shape of the imprint chara- 
cter is better for developing mathematical symbols. 

Dr. Nemeth : May I make a comment about that. There is so much 
science around now in mathematics that no transcriber can be expect- 
ed to have a comprehension of everything he's asked to transcribe. 
Consequently, unless the transcriber can transcribe what he sees, 
the likelihood is you get no transcription whatever. In other words, 
the ideal thing would be if someone could convey to you the meaning 
of the thing instead of the contour of the thing. But it's just 
not possible. There are no transcribers that you can get who will 
understand algebra, atomic physics, statistics, and what have you. 
Therefore, the only safe way to do is to convey notations instead 
of significance, and let the blind person read the significance 
out of the notation, I'm firmly convinced of this. 


Mr . Inghsun ; We're not arguing the Nemeth Code, 

Dr. Nemeth ; No. We're arguing the philosophy of presentation. For 
example, let me go to the English braille and I'll tell you some of 
the things that arise in my mind where this factor has been neglected. 
One of the rules of the English braille system says that you should 
put abbreviations in front of the number sign instead of after the 
number sign. If you want to write "50 feet" you write "ft #50". 
This conveys the meaning perfectly well, but you know that a blind 
person grows up with a misconception of what that imprint it. He's 
likely to do the very same thing on a typewriter and get a grade 
for using bad grammar from his teacher who doesn't know what the 
rules of braille are. In other words, writing "ft #50" is the kind 
of thing that I call a braillism. It conveys the meaning perfectly 
well but it does so at the expense of the blind person's full com- 
prehension of what the standard situation is in inkprint. 

Another thing: A rule of braille says that a transcriber should 
translate Roman numerals into Arabic numerals even in references and 
so on. So if you want to write II Samuel (Roman numerals), you don't 
write the Roman numeral II but you write the Arabic 2. It has come 
to the point where I have to have a file card available for my re- 
ference as to what the proper way is for writing a reference, because 
if I were to use the rule of English braille, I would be accused of 
being an illiterate professor. No one writes references like this 
in print. And I very often have to use the typewriter. You see, the 
blind person's life doesn't begin and end with braille. The inform- 
ation has got to come to him from something which has previously been 
written and often he has to convey information in written form. So 
he has to be aware of both of these, as to what normal print practice 
is if he is not to labor under a misconception as to what is right 
and what is wrong. The case in point is that I have to keep a re- 
ference card as to what the right way is to put a reference in a 
footnote because the way you do it in braille is just wrong. You just 
don't do it that way. This is the point. 

Mr. Covici ; Fine. This is a good point, but is it so bad? I* mean 
can a person remember that it's done differently in print? 

Dr. Nemeth ; He never gets a chance to find out. 


(Dr. Nemeth and Dr. Foulke then proposed some ways and means 
of determining the intrinsic degree of ambiguity of certain six 
and nine-dot braille characters relative to code design which 
were discussed as follows.) 

Dr. Nemeth : I haven't done this, but this is a perfectly mathematically 
feasible thing to do. Let's go to the six-dot braille cell and let's 
imagine that we have before us a sheet of paper of infinite density 
so that it is high and wide and we put braille characters on it. Some 
of them are truly unambiguous; for example, the "y" or the "sh" 
sign or the full cell, but there are certain other cells which have 
certain degrees of ambiguity. For example, the letter "a", if it 
were present or by itself, you might say has six degrees of ambiguity 
because if it were present all by itself you could not tell which of 
the six dots it was. The letter "b", you might say has four degrees 
of ambiguity because it could be 1-2, it could be 2-3, it could be 
4-5, or it could be 5-6. Similarly, you can assign an uncertain 
factor to each one of the 6A- braille characters. The space is a . 
perfectly certain character, because if there is nothing on the 
paper at all, then evidently a space has been printed. The only 
question is where! 

You can do this and you can sort of weight the amount of uncer- 
tainty that there is in standard braille by multiplying each one of 
the characters by the uncertainty factor which you have associated with 
it, and then you can divide the result by 64 to see how much uncertainty 
there is. 

You can do a similar process with a "Nine-Dot Braille" cell. You 
see, when you read braille the uncertainty comes out because of contig- 
uity. Things are next to other braille characters or out of context. 
For example, a dropped "h" can be either a question mark or an open- 
ing quotation mark, depending upon whether it is at the end or the 
beginning of a word with which it is associated. 

Mr. Covici ; That's a six-dot. 

Dr. Nemeth ; That's right. The point it — 

Mr. Strom ; Only one side has one meaning. 

Mr . Ingham ; Dr. Foulke »s suggestion is that we are talking about 
other codes and all of these points should be kept in mind in designing 


Dr. Nemeth ; That's right. But what I'm saying is, this is a way 
of assessing mathematically the intrinsic uncertainty of ordinary 
braille versus the intrinsic uncertainty of "Nine-Dot Braille". 
Does everybody understand what I'm talking about? 

Mr. Strom ; I understand. 

Mr. Covici ; It may be overweighed, overridden by a lot of other 
things, you know. 

Dr. Nemeth; This may be true, but I'm talking about intrinsic un- 

Dr. Foulke ; This can be determined. 

Chairman Rodgers ; Hold it. We're getting into cross conversation 

Mr. Roberts : Both Mr. Strom and Dr. Foulke wanted to react. 

Chairman Rodgers ; I think Dr. Foulke spoke ahead of Mr. Strom. 

Dr. Foulke ; I think this could quite easily be done. If I were 
going to conduct research of this sort, for instance, I think one of 
the things that I would quite likely do would be to give people se- 
quences of symbols to read and to cast their results in a confusion 
matrix to determine such things as stimulus ambiguity. The procedures 
for doing this are quite well known. 

Dr. Nemeth ; Yes. 

Dr. Foulke ; I think I would go ahead and make the standard inform- 
ation analysis. 

Chairman Rodgers ; I think Mr, Strom spoke next, or tried to. 

Mr. Strom ; Yes. One thing that has to be mentioned. I understand 
what Dr. Nemeth had to say, this numerical measure of intrinsic po- 
tential for confusion is a function of number of dots alone. Is 
that it? 

Dr. Nemeth ;. No, For example, suppose you decide to have nine dots 
around the circumference of a circle instead of ranged in a square. 
Then the nine dots in the circle would be the same as the number of 
nine dots in the square. So it wouldn't be a function of the number 


Mr. Strom ; I'm sorry. That's right. But it is a function that's 
independent of the assignment of symbolic meanings? 

Dr. Nemeth : That's right. I have no intention here of finding 
out what meanings you assign to the letters. It's just a measure 
of what I call the intrinsic ambiguity of the system - that's all. 

Mr. Strom ; Right. Although a more interesting measure, I would 
think, would be to take six-dot braille characters and observe with 
respect to each six-dot braille shape or configuration how many 
different meanings are assigned in that system to that particular 
configuration, such as — 

Dr. Nemeth ; Take the letter "n". What do you mean now? 

Mr. Strom ; Take the letter "n". It has two meanings. It means 
the letter "n" and the whole word "not". 

Dr. Nemeth ; "not" and that's all. 

Mr. Strom ; Okay. 

Chairman Rodgers ; Take the dropped "d". I think it has four 
meanings . 

Dr. Nemeth ; I know it has several. I think this is true. 

Mr. Strom ; Then what you do is you weigh these up and you divide 
by 6^. 

Dr. Nemeth ; Not by 6^. 

Mr. Strom ; I'm sorry. By the total number of meanings that there 
are . 

Dr. Nemeth ; No, no. You see when you do this you're measuring what 
we call contextual uncertainty and then you have to make a statistical 
analysis of the trequencjf with which, say, the period occurs, the 
frequency with which the "dis" occurs, the frequency with which "dd" 
occurs in the English language you know, from large samples, and you 
have to weight this. With this kind of a frequency distribution it 
would not be fair to say that, for example, the accent mark occurs 
as frequently as the letter "e" and use it as weighting the things 
equally, you see. 


Mr. Strom ; Yes. 

Dr. Nemeth ; You have to make a distribution of the frequency of 

Mr. Strom ; You are right, and I think this distribution should be 
made and the results of this study would be a little more significant 
and would be a measure of the so-called intrinsic ambiguity in "Nine- 
Dot Braille", which is effectively presupposing that because the 
nine-dot cell is bigger, it's possible for the letter "g", for 
example, to have one, two, three, four different possible positions. 
This will, therefore, tend to increase the ambiguity coefficient. 

Dr. Nemeth ; On the other hand, there are so many characters which 
are fully wide and fully long, more so than in the six-dot braille, 
that the factor will have the effect of decreasing it. 

(The discussion on braille-code design then centered on the possibility 
of exploiting the full potential of "Nine-Dot Braille", the develop- 
ment of experimental codes and the possibility of restricting the 
use of "Nine-Dot Braille" to special technical codes only). 

Dr. Ashcroft ; I don't know if this is relevant or not, but it seems 
to me that some of the kinds of statistical computations around 
which questions have been raised today could be attacked right away, 
such as the amount of space and the comparative space-saving features 
of Mr. Strom's system, and that could proceed right away, and then 
I think we could map a strategy, at least a tentative one, right 
here for attacking some of the experimental work for a tryout of 
some nine dot systems. 

Chairman Rodgers ; Of nine dot cell characters; is that it? 

Dr. Ashcroft ; Yes. 

Mr. Strom ; This would be contingent on developing a code. 

Mr. Ingham ; I don't know how you do that unless you're assuming 
this code. 

Dr. Ashcroft ; Let's assume this code for some tentative answers 
that we haven't had today. 

Mr. Strom: Okay, 


Mr. Ingham ; Haven't you gotten that already, any of this? I mean 
your figure of M-0 per cent on the math and things like that? 

Mr. Strom ; I have not made exact measurements. First of all, this 
code and the code of two years ago are different. Furthermore, I 
don't really consider either of them representative of the full 
potentialities of "Nine-Dot Braille". I think these codes both 
represent an extreme construction of the possibilities of "Nine- 
Dot Braille", namely that they have been harnessed to certain ideas 
which are part of the six-dot concept, namely, the concept that 
every contraction must be an initial or final-letter contraction and 
must contain a base cell and a modifier in a certain way,* 

Also this code has been based upon a principle that anything in 
six-dot braille like punctuation marks, and so on, should not be 
modified unless we are sure that there is something wrong with it. 
So I don't think these two codes are representative, and although 
I would be perfectly willing to have various kinds of tests regarding 
spacing, and so on, set up with these, I would put in an element of 
caution and say that this particular code would not be truly rep- 
resentative of what "Nine-Dot Braille" really could be if we gave 
a little bit freer rein in designing a code. 

Chairman Rodgers ; All right. 

Mr. Ljechty ; A question that's been bothering me, Carl, that I 
would like an answer to: I don't know whether it's the right place 
or not to ask, but I'm going to do so anyhow. 

What bothers me is that it is naive, possible to ask: Could 
not a technical code be devised on the basis of a nine-dot cell, 
that would fit in mechanically as a unit with the standard six-dot 
cell, so that you could write a mathematics text using the narrative, 
the words in literary braille, making use of the standard literary 
braille and simply insert in the line, when the occasion occurs, the 
technical sign based on the nine-dot cell? That would permit inter- 
pointing of dots on both sides of the paper because your spacing 
would be standard and uniform. 

Mr. Ingham ; Could I ask for clarification. What do you mean? Do 
you mean that, for example, within an equation where you had, for 
example, "x" equals "y" square, that "x" would be standard six-dot? 

Mr. Liechty ; Yes. 

Mr. Ingham ; The equals and others would be nine-dot? 

"This is an incorrect description of six-dot Grade Two braille design, sinc« 
there are sixty-seven one-celled contractions, consisting of base cell 
alone, while there are forty-seven two-celled contractions consisting 
of base sign preceded by modifiers, 


Mr . Liechty ; Right. That's what I would think, but I don't know. 

Mr. Strom ; Let's consider something a little more realistic; 
namely, where a line of a formula is all in mathematical braille, 
considering that there is an "x" for mathematical braille also. 

Mr. Ingham : The reason I brought that up, then, is that if 
you're going to go with a nine-dot system for your equations and 
formulae, and we're not going to say that the existing English code 
is the correct code for "Nine-Dot Braille", then I don't know how 
you could represent your letters and things like that, and numbers, 
in the two different systems and expect people to be able to read 
them or want to read them. 

Mr. Liechty ; This is my question. I don't know with codes. So I 
don't know how 'to answer, Carl, may I make one more comment and I 
think this is the most important comment that I have to make all day; 
I'm not so sanguine to think that within our lifetime, or in the 
next two or three generations, a basic change from the six-dot to 
a nine-dot braille cell is going to be achieved, if it is going to 
have to apply to literary braille. 

Dr. Scott ; Mr. Rodgers, I wonder if we could ask Mr. Liechty for 
clarification of that. Why isn't it likely to happen? 

Mr. Liechty ; In my opinion, and this is only an opinion, there 
are enough contractions for the average blind reader to learn and 
use now. Anymore will defeat any further reading in braille for 
the vast majority of people, and this is why I first asked the 
question whether or not we can make a distinction and use a new 
basic cell for only technical purposes. You will not get John Jones 
of Podunk, Iowa, to drop his six-dot cell and go to a nine-dot cell 
and learn a new system of braille, 

Mr. Ingham ; " I think you are probably quite right in the case of 
John Jones, but if you ask me or anyone else in the scientific 
fields or music, I might say to you I would be willing to know two 
different codes, six-dot and nine-dot. 

Mr. Liechty ; I'm sure you're right. My interest in "Nine-Dot 
Braille" has been on the basis that this will serve a special people, 
an especially selected group of people. It will serve technical, 
scientific, mathematical purposes but not for literary braille, 
for novels. To read a novel in another system of braille will not 
be achieved. 


Mr. Covici ; Would it be worth it to you if it increased the speed, 
say, by a third? I'm just throwing out a figure of a third, 

Mr. Ingham ; I don't think it's a question of that John. I think 
it actually will happen, but he's right, it will take two or three 
generations for the average guy reading literary braille to move 
over, but so what? 

Mr. Liechty : In that period of time you would have a duplication 
of systems and libraries, and confusion. 

Mr. Inghsim ; No duplication of the scientific libraries. 

Mr. Liechty : No, but in the literary world.' 

Mr. Ingham ; Yes. This may be worth it in the long run. You may 
find that some such code turns out to be really worthwhile and to 
have a distinct advantage, so that it may take over the literary 

Mr. Liechty : I think here then there is a very proper place for 
research to find out whether one can determine what is the top limit 
of the number of contractions that will be assimilated by the 
reader, the blind reader. 

Dr. Nolan : I think the value goes beyond that. Maybe nine dot 
isn't the answer, but by studying "Nine-Dot Braille" maybe we can 
learn more about tactual reading and devise some other system 
that may solve our problem. 

» * « « i» « ]^ 



Chairman Rodgers ; Does anyone have any question relevant to the 
other pror)osed items of discussion? 

Mr. Covici ; I have one question in respect to small and large 
quantity duplication. 

Chairman Rodgers ; I think you are referring to the point of de- 
termining whether small and large quantity duplication would be 
feasible with respect to "Nine-Dot Braille". By small quantity 
duplication I meant such things as braille ^writers, braille slates, 
and so on. The reason I brought that up, (see Appendix A Working 
Paper), is that I think you will recall, Mr. Liechty, at the last 
conference, some people were concerned with that aspect, 

Mr. Liechty ; Yes, 

Chairman Rodgers ; Certainly, duplication of "Nine-Dot Braille" 
should be studied in terras of the big IBM computers. But we should 
also include the study of duplication by the conventional methods 
of stereo,typing equipment, braille writers, and braille slates. 

Mr, Liechty ; That's an ordinary mechanical problem, 

Mr. Covici ; I think that research on duplication should be done 
along with research on code development. 

Mr. Liechty ; Still, Carl, you would have the mechanical adaptation 
to make of the embossing machine. 

Mr. Covici ; Yes, adaptation of duplication techniques and code 
development should proceed concurrently. 

Chairman Rodgers ; That's a mechanical and engineering problem and 
it certainly should be considered. 

Mr. Strom ; Mr. Ingham, would you know anything about the possibility 
of getting machinery produced so that we could get more of this 
material made? I mean I'm thinking of the problem that at the moment 
I have a two-line slate. I was wondering whether we could recommend 
that we look into the possibilities of getting a machine to do this? 

Mr. Ingham ; There is a possibility in this sense that if you could 
design a program which would produce the "Nine-Dot Braille" following 
your rules from a typist through, say, the IBM machines, it's very 
possible that the new MIT high speed brailler could be adapted, I'm 
not sure how the keying works, but it probably could be adapted to 
produce a heavy-weight paper sample which could then be thermoformed. 
The cost there simply would be the cost of the therraoform. 


Mr. Strom ; I would like to set up an arrangement whereby this 
could be done. 

Dr. Nemeth ; Do you want all the same material distributed or do 
you want lots of different kinds of material distributed? If you 
want fifty pages of the same material, it's not unreasonable to ask 
somebody to punch this out on a slate. Don't laugh. Volunteers 
d'o this all the time. They transcribe volumes at a time. If you 
want to do fifty pages of work, you can do it. I mean somebody 
could punch it out in fifteen hours of gross time. Then you could 
thermoform it and that could solve the whole problem. 

Mr. Ingham ; I know we're all mechanically oriented, but I think 
this is a simple solution. 

Mr. Strom ; Yes, but there would be errors. For example, I made 
22 pages of material for this conference which never were thermo- 
formed because of the reading difficulties that we encountered, but 
Mr. Rodgers and I looked over it, and there were many erasures that 
had to be made. 

Mr. Ingham : You have to have an expert transcriber. These are 
available . 

Dr. Benham- : May I make a point, please. At Haver ford I built a 
machine a few years ago that is like the Perkins braille transcriber, 
the big thing, except that it's solenoid operated. It's relatively 
crude. I made it for the purpose of making legends on diagrams and 
things of that sort. It writes on zinc plates. I think for maybe 
four or five hundred dollars an extra row of holes could be added 
to it to make it nine-dot. Then you could write on a metal plate 
with it. 

Dr. Nemeth ; What about the escapement on it between cell and cell? 
You would have to vary it, you know; you would have to make it bigger. 

Dr. Benham ; You can make a new rack for that. That's all it is, 
just a rack. 

Dr. Nemeth : All right. 

Chairman Rodgers ; Excuse me. I think we have become too involved 
in the point of mechanical reproduction. I would like to ask Dr. 
Nolan — I know no one can commit anybody to anything — but, Dr. Nolan, 
do you think the American Printing House for the Blind might be in 
a position to help with the reproduction of text materials? 


Dr. Nolan ; That would depend on several things: One, what we 
make them from. Are you talking about vacuum forming? We can 
vacuum form all the stuff you want. If you're talking about build- 
ing machines to make "Nine-Dot Braille", this is a costly enter- 
prise and there might be some questions attached to it. 

Mr. Ingham : Can I offer another suggestion. There's another 
machine we have in the lab which is what we call a picture brailler 
which loads the heavy-weight braille sheets on a rotating drum and 
can emboss a 90" x 90" array of dots from recorder or from a com- 
puter or from what have you. I'm sure that, for this kind of 
density, we could adapt it with a simple program again to do a 
master for thermoforming of this type of material. After all, 
there are plenty of graduate students and senior thesis students 
who could do such a prograim and set up such a thing. 

Mr. Strom : Given a description of the machine, I could write the 

Mr. Ingham : Fine. I mean this is a simple solution. 

Dr. Ashcroft ; It seems that multilith-multigraph approach might 
be very easily a solution. 

Mr. Covici : There's actually a machine into which you put a 
paper tape, and it will make something. 

Mr. Strom ; It will operate a tool. Anything that involves progr 
ming. I don't know that particular machine. 

Mr. Ingham : I was thinking of the flexowriters and machines of that 

Mr. Strom ; They have machines to control the operation of the tools 
from a program. 

Mr. Ingham ; Yes. 

Mr. Strom ; So that you could, for example, if you wanted to make 
a nine-dot slate with a certain specification of dot spacing, you 
could set these specifications up on a paper tape and the paper 
tape would control the operation of the drill. That could make a 
slate. Again, before I could work with this, I would have to do 
things like learning the machine language for that machine paper 



Mr. Ingham ; There are minor things, 
Mr. Covici; Yes. I just mentioned it 




Co-ordinator 's note : At various times during the Conference, the 
participants posed questions with respect to the role the Foundation 
might be prepared to play in carrying out the participants' recommend- 
ations for research on six and "Nine-Dot Braille". 

In the interest of conciseness Mr. Roberts' replies have been 
consolidated as follows: 

Mr. Roberts ; Certainly the Foundation is very much interested in 
furthering any efforts that will result in improved systems of 
reading for blind people, and I think principally what I would like 
to see come out of this is your advice really as to how we can furt^^er 
this mutual ob jective . . . . .If there is a role that we have to play and 
can play that would De useful, why, as Mr. Barnett said, we certainly 
would use our resources to the best of our ability to play that role, 
and it is hoped, effectively. If this is to be done alone by the 
Foundation, if it is to be done by others outside of the Foundation, 
if this is to be done collaboratively, we have no predetermin.ition 
on this at this point. We are really looking for your advice and 

I should now like to pose a question: What apparatus is there 
in this field for launching any kind of organized system of research 
that can be identified as a valid proposal here? I should add ( and 
this is what I implied before in a more direct response to one question 
that was asked) we have no plans for carrying out research in this 
area at this point; we have none whatsoever. It may be that we will 
and it may be that we won't. I think we want your advice about this. 
Let us assume that we identify and agree on areas of research that 
are sound and valid. The question I think needs to be thrown out to 
the floor as to what the auspices of this should be; what the possible 
resources are to undertake this research. We don't know the answer 
to this now..., If any proposals come out of the deliberations here 
that would seem to indicate a need for this organization--and it does 
not necessarily have to be the case — using its resources (dollars, 
equipment, staff) to carry out the proposals, we certainly would con- 
sider this very sympathetically. 

If I just may make this one other comment. I think this has 
been implied in my remarks; maybe I ought to state it more directly. 
That is, that we have quite frankly no sense of paramount interest or 
concern or responsibility in this whole subject that we are talking 
about. I think it has been perhaps a long time since this organization, 
itself, has conducted research in this area, if it had at all in the 
past. I don't even know that,,.,. At this point, I should add, we 
have no plah to ask any particular individual or group to undertake 
research. It hasn't as yet really been agreed to or identified in 
a formal sense. 


Chairman Podgers ; Now, I think we are ready to attack the strategy; 
in other words, to determine what method of attack we're going to 
use to develop a well structured long-range research project. I 
think it was Dr. Ashcroft that used the word "strategy", did you not? 
And for that reason I would like to start the discussion by asking 
Dr. Ashcroft for his ideas on strategy. 

Dr. Ashcroft ; It seems to me that we need some kind of a place to 
start, whether it's with Mr. Strom's present version of nine-dot or 
some modification of it and to get samples of that out for reading, 
and to include at that time some of the options that come up to get 
these trial readers to feed back options that they see developing for, 
say, the assignment of meaning to symbols or introducing more system 
to the assignment of meanings. Then to test some of those options in 
research and feed the answers back into a revision of the original 
version and then we would be ready, it seems to me, for a fairly ex- 
tensive trial, experimental trial of the system. 

Mr. Strom : Who would operate this? I mean how would such a test 
procedure be organized? 

Dr. Ashcroft : I think the possibility is limitless there. I think 
if you could just get information out about the problem and the idea, 
that you will have competent people respond to the challenge, and I 
think funds can be obtained to do it, however they would plan to do 

Dr. Foulke : If we can tell them what to do, they will do it. 

Dr. Nolan : I would not phrase it exactly that way! 

Chairman Rodgers ; All right. Dr. Nolan. 

Dr. Nolan : I think with the thought I have given to it, what I 
might do would be to start out on this legibility thing. I'm so con- 
cerned over this. I think Mr. Ingham differs. Maybe some of the 
others of you do. Let me tell you why I'm concerned. 

In order to get the sensation of touch, the surface of your skin 
has to be, for want of a better term, dented. There has to be a dent 
made in it. With our regular six-dot cell there's enough space on 
either side of the cell so that the skin can droop down, if you will 
pardon the use of another non-scientific term, and we can get this 

When we go to a nine-dot cell using the same inter-dot spacing 
with three dots in a row, the closeness of these is not going to 
allow for an indentation to be made by the center dot and this is 
going to be lost to many readers. 










Dr. Benham : Is that any more true in a horizontal row than in a 
vertical row, which we now have? 

Dr. Nolan : I don't know. It ought not to be, but you see what 
we're running into. We're not actually just talking about rows; 
we're going to have rather gross patterns, are we not? We could 
have nine dots or we could have eight dots in many figures or even 
seven dots. The point of this is until we're sure of the legibility 
of these things, to start researching codes may give us problems 
because our problem may not be with the code, it may be with the 
legibility of the configurations. 

Dr. Ashcroft: We have tentative evidence of the legibility already. 

Yes I think so. 

What do you mean, tentative evidence? 

We can read it. 

The very reason that you're here is because you're at 
a higher level of achievement than many of the people who are going 
to be using this thing. 

We need to determine legibility for the full range of ability 
found among braille readers. 

Dr. Foulk e : I don't think we have any kind of evidence of legibility 
that we would be willing to act on, do we? 

Chairman Rodgers : I think it was Dr. Ashcroft, or was it Mr. Ingham, 
who mentioned that we have tentative evidence. 

Dr. Ashcroft : I did. 

Chairman Rodgers : I would like to hear it Dr. Ashcroft. Will you 
put it on the record. 

Dr. Ashcroft : One of the earliest comments I heard at this conference 
was that even the crude material that you had last time, which had 
spacing problems in it and other problems, was read by people here. 

Mr. Ingham : I think this is probably derived from my comments. I 
mean I'm no protagonist for "Nine-Dot. Braille" ; I'm willing to go 
either way. I still think we should do research. 

Dr. Ashcroft ; I think there are two approaches to the problem of 
legibility. One is a gross one where you just provide people some 
stuff and let them see if they can read it, or the analytic one that 
Dr. Nolan is suggesting, and I would certainly go to the gross one 
before the analytical one. 


Dr. Foulke : I don't think I would. If I were going to start re- 
searching this problem, I would conduct, I think, an initial study 
to try to determine or learn something about the legibility of all 
of the characters that were possible in this system, and I think I 
would probably want to set up at least some kind of tentative criteria 
for symbol legibility, and on the basis of this then attempt to form- 
ulate some kind of code. 

Dr. Ashcroft : Then you will have to do the legibility studies on 
symbols in isolation and you won't know the answers to whether it 
can be read or not. 

Dr. Nolan : Wait a minute now. Let's just go a bit further. I think 
a check on legibility should certainly be made. This doesn't mean 
that everything has to stop while this is being done. It may be that 
the initial efforts on legibility will show that this is really an un- 
necessary problem. It may be that we will see that we really have 
legibility problems and if this is the case, then it's rather reck- 
less at the moment to get involved in codes until we solve this problem. 

Mr, Covici ; What do you mean by legibility, counting how many dots 
there are or recognizing the difference? 

Dr. Nolan : There might be a very elementary way. You might dis- 
criminate different patterns or you might just report the stimuli 
present. There are several approaches that you might use. 

Chairman Rodgers : Just a moment. I wonder whether the two phases, 
the legibility phase and the stimulus phase, can be investigated con- 

Dr. Nolan : What I think I would do if I had this problem to solve, I 
would work on legibility. I would start exploratory work with the 
code Mr. Strom has devised. This would be a trial and error process — 
actually trying to accumulate sources of error in the code. While I 
was doing this also I would be refining the code on a logical basis 
and also working into these special areas where special codes need 
to be developed. Then at the same time, or a little later, I would 
just for the heck of it develop a whole new literary code and study 
this to see how we rairht alter the degree to which we can communicate 
in tactual form. 

Dr, Ashcrof t: By new literary code you mean six-dot or nine-dot? 
Dr, Nolan : Nine-dot. 

Dr. Ashcroft : How would you differentiate what we've been talking 
about from literary code? 

Dr. Nolan : The present setup is to use the old literary code as the 
base and modify it, isn't it? 


Mr. InRham : That's right. 
Chairman Rodgers : Yes. 

Dr. Nolan ; We know, on the basis of our experience with the present 
literary code, somethinr^ about soft spots in it. For example, fre- 
quency of occurrence of letters, etc., are not reflected in the pre- 
sent code. I would develop a new literary code on the basis of what 
we know about information theory, what we know about the problems we 
have found in your present literary code, and so on. Remember, vre're 
researching, we're studying, we're not saying that we're going to 
adopt any of these things, that any of them will ever c-me into prac- 
tice. But where are we now? Here we've had this problem for years. 
We have "trialed-and-errored" it through. No one has ever extensively 
explored beyond the present system; what has occurred has been more 
a fight and a power struggle than an exploration. Here we have an 
opnortunity to take off in several directions at the horizon. What's 
over it, we don't know. Let's look and see! 

Dr. Benham ; It seems to me that there are few enough people reading 
braille as it is, and the harder we make it to read the fewer there 
will be. I don't know what the percentage drop in reading was when 
it went from Grade One to One-and-a-half and then when it went from 
One-and-a-half to Two, but I'll bet it drops every time there's an 
increase in the complexity of it. So I think that Dr. Nolan's 
suggestion of finding out the legibility of the system, nine-dot 
system or any other system, is extremely important because, yes, we can 
read this. 

I agree that at the first conference we had, I could read the 
"Nine-Dot Braille" that I was presented with and, in fact, I believe 
I even wrote a little of it on the slates. Sure, I could do it. But 
it was a contest; it was a game. I was vying with my fellow man 
trying to be as clever as he was and he was trying to be cleverer than 
I was,. and as a result we were stimulated sufficiently to manage to 
struggle thought it. But I imagine I'm the poorest braille reader 
here today. Because I can read braille only when it's absolutely 
necessary that I read it, like for instance, the Working Paper that 
we were presented with. I read through those pages, but it was a 
struggle. I don't know why this is so; I wish I did. But I really 
get frustrated trying to read braille. I don't know why it is. It's 
obviously some psychological thing in my background some place, but 
that happens to me, 

I can struggle through mathematics and physics and engineering 
and science of all kinds of that nature because it's a slow, tedious 
process reading it anyhow. So it doesn't bother me. But to pick up 


braille and read it as a novel, I just wouldn't think of it, I 
couldn't. I can't even read the braille technical press (and I'm 
interested in what's in there), partly because there is so much re- 
dundancy in it. In your working paper, for instance, I nearly 
through the thing out the window when I got to the last couple of 
pages and you repeated and repeated and repeated in each line. Why 
you couldn't have said it at the top of the page and then given the 
salient features of what it was you were doing, but every paragraph 
is just like every other paragraph except there was one number 
difference in it. And I had to read through the whole thing just 
to find that different number, and I was afraid to skip it for 
fear you might have slipped something in on me. 

Chairman Rodgers : You're now talking about the format of the 
braille; is that it? 

Dr. Benham : No, I'm just talking about the fact that it's hard 
for me to read; it's slow, tedious, and I read it only when I have 
to and, of course, I have to, and I have to quite a lot. But I don't 
do it for fun. And the Grade Two contractions throw me every once 
in a while. I'm trying to read along and I come to some cockeyed 
looking this. I can't make out whether it's capital "St" or "each" 
or some other combination of dots, I'm not being very coherent, 
but then I'm not intending to be. 

The point that I'm trying to make is that I think that when 
we go to something that is harder to read than what we already have, 
we'd better be sure that it's going to be acceptable to, I would say, 
at least as many people as now find Grade Two acceptable, because 
we're going to lose people on the way. Another interesting comment 
that I have often though of is: I wonder whether blind people are 
poor readers because they never see words spelled. Everything 
you read in braille is a contraction of some sort as the "m" for 
"more", but you never see "more" spelled out. We're going to get the 
system so contracted that its going to get even worse in the literary 
area if we're not careful. I think I told you at the beginning that 
there was a great deal of redundancy, because I agree with a great 
deal of what's been said, I was only trying to emphasize a few 
points. For literary braille, I'm not sure that we want to make the 
thing more contracted than it is, more difficult to learn, and more 
difficult for some people at least to read. We ought to try to make 
it easier to read, not harder, 

I find also that the "Nine-Dot Braille" that I've seen so far 
is too spread out. For instance, when you take the material that 
you presented us with today, the code, and you look at the nine-dot 
version of it and then look at what it means, like "pro" or whatever 
the letters are, those letters look like they were sitting in the 
middle of the desert, 


Mr. Covici ; You mean the six-dot braille. 

Dr. Benham ; Yes. But obviously you're going to have to spell 
some things out. You're not going to have everything abbreviated, 
I hope. So there will be occasions where you want to follow one 
letter with another. What is it going to look like when you do 
that? Is it going to be spread out? Now, I've been inconsistent 
because in some cases I say you're getting it too compact, and in 
some cases I say you're getting it too spread out, I think it 
boils down to the fact that I believe the present braille spacing 
in the six-dot code is about the optimum spacing for easy reading. 
To get it much closer, you have to study it too much except for 
the exceptional braille readers, and if it's too spread out it gets 
difficult to read and you can get lost in it. This is one of the 
things that I dislike about the Nemeth Code. It's too spread out 
on the page as far as I'm concerned. Of course, I was brought up 
on a different one, so my opinion is really not important, but 
for me it's too s-oread out in many cases. I want to find out what 
the legibility of a new system is, before we go very far in trying 
to decide what kind of a code it ought to be and what kind of ab- 
breviations there ought to be. 

Dr. Ashcroft : I was much in accord with what Dr, Nolan said about 
carrying on legibility studies, but I think some kind of actual 
reading study, some kind of a nine-dot code, should go parallel with 

Dr, Nolan ; Yes, 

Mr. Ingham ; I agree. 

Mr. Covici ; Yes, 

Dr, Ashcroft ; I think this is in line with what Dr. Benham has 
just said, too. 

Chairman Rodgers ; Yes, I think so, and perhaps the two studies 
could go on concurrently, but I would like to comment in particular 
about what Dr. Benham has said with respect to reading difficulties, 
I think that in the study of legibility, one thing we have to be 
very careful about is not just whether the finger can discriminate, 
but how readily? In other words, there is a difference between 
mere discrimination and getting by, and genuine reading; and that 
is, I think, what we're concerned with, 

Mr, Covici; It would take training. 


Chairman Rodgers : Yes, it would take training. In other words, 
one should be able to discriminate without too much study of the 
thing, you see. That's the point I think that should be very care- 
fully taken into account in any study of legibility. Maybe I mis- 
used the words "degree of discriminability" . I should perhaps have 
said "discriminating without too much studying of what you have 
before you", 

Mr, Strom ; Then, Mr. Rodgers, would you say that the experiment 
should go along the lines of a point where we spend several months 
training a certain number of subjects to become proficient with 
"Nine-Dot Braille", and then make measurements with respect to how 
quickly they pick them up? 

Chairman Rodgers : Yes. I just tried to throw that in about ready 
discrimination insofar as helping the study is concerned. But as 
to what specifically to do, that's a matter of structuring and that 
is what we're here for, to collect your opinions as research special- 
ists. That should be up to you to decide and if anyone would care 
to comment on Mr. Strom's suggestion — 

Mr. Ingham : What was the suggestion again? 

Mr. Strom : That among the tasks that we consider would be one in 
which subjects would be trained for a few months to learn the code 
and develop proficiency in adapting to its peculiarities, and then 
train them on being able to read straight text of the material and 
see how easily they are able to, 

Mr. Ingham ; Which code are you talking about? 

Mr. Strom ; Either the current "Nine-Dot Braille" code or one of a 
few modifications thereof. I think the point mentioned about two 
"Nine-Dot Brailles", namely, one like the one here and another one 
that we invent on the basis of six-dot braille but, rather, on the 
basis of information theory, optimization of whatever factors we 
figure will make it more psychologically readable. I think we should 
try, perhaps, working with both of these, but the important thing 
is that we have got to get some kind of tests made with some code. 
The more codes we get tested, perhaps the better off we are. 

Mr . Ingham : What Dr. Nolan was saying requires that you people 
go into lots of practice. If you're going to use a new one, you 
have to have lots of practice, of course, i 

Mr, Strom ; Then we have to first of all get people in our project 
who are specializing in teaching braille, I mean among the people 
who would carry out a research program will be specialists in 
education, people who are familiar with the specific problems involved 


in teaching of braille and who would be able to adapt this to 
"Nine-Dot Braille" so that if we have a research group, one of the 
members of the group would include a specialist in education. 

Dr. Scott ; Have you, Mr. Strom, at the present time, developed 
any code which exploits to its maximum the potential "Nine-Dot 

Mr. Strom ; No. I haven't. 

Dr. Scott ; Would this be a relatively simple thing to do or 
would it be time consuming? 

Mr. Strom : It would be relatively easy to invent a code that 
broke away from the six-dot braille and gave more flexibility. 
What would not be relatively easy would be to optimize certain 
details within this. 

Mr. Ingham : You actually could, particularly if you could take 

advantage of, say, the lists of the ^00 more frequent words in 

English text, and so on and so forth, which the English braille 
could do. 

Mr. Strom : Yes, but also what I have to take into consideration 
are the 512 nine-dot configurations and some kind of measure of 
the degree of appropriateness of these contractions. So you have 
a whole bunch of lists of common words or word parts and then you 
have a list of 512 dot configurations. We have to somehow isolate 
a property of these orthographic configurations which we can relate 
to a rule for assigning proper meanings to them. 

Dr. Foulke : This you can probably do on the basis of linguistic 
information, for instance, construct a code taking into account 
such things as sequential dependencies, orthographic conventions, 
things of this sort, but in addition to that, before constructing 
a complete rational code, it seems to me that you would have to 
know something more about the stimulus alt)habet itself and we're 
back to legibility again. 

Mr. Strom : That's right, 

Mr, Govici : Can't you do both things at once? 

Mr, Strom : The problem is I could write a tentative code but I 
could not write easily an optimum tentative code. What would have 
to happen is first an intuitive code and then give it a legibility 
test; train subjects to read it, assess their reactions, and get 


verbal questionnaires from them, get statistics on speed and, in 
short, get quantitative and qualitative data on this. Then have 
the specialized studies regarding problems of individual characters 
independent of meaning, as somebody suggested. Then we have these 
two studies. These have to be correlated. 

Then there has to be a feedback mechanism where the code is 
revised and -out through the same tests, so that you are always 
using information that you are gaining about psychological prop- 
erties, legibility of characters relative to the meanings you 
assign to them, and you have to use this information and put it 
in the context of a whole system which you then have to put to the 
tests. The tests will take a long time. Every time you have a 
new code you have to get some new subjects and train them for 
several months on this. Presumably these tests would have to be 
made — you would have to get a whole group of blind subjects. 

Mr. Covici : Maybe we could have the subjects optimize the code 
by these tests, maybe change it a little bit and make it better 
and then eventually we can get enough legibility studies through 
and then maybe eventually we will have one code. Then we will try 
it on a whole group of people and get your statistics back. 

Chairman Rodgers ; It seems to me that that is a statement that 
the research specialists would have to approve; that is, in terms 
of established norms of research procedures. Whether it would be 
a sound one would be up to the research specialists to determine. 
Would any one of you care to comment on Mr. Covici 's remarks? 

Mr. Ingham : I think it's haphazard. That's no way to go about 
doing it. Could I interject a hopeful note as far as one of the 
things that was brought up. One of the hardest things I had to do 
was to find how many people do read braille. If anybody has a 
concrete fig:ure, I have, and this is somewhere between 30,000 
and 40,000 who are reading braille at the present time; this number 
is constantly increasing and has never declined actually. 

Dr. Nemeth : What constitutes reading braille? What material do 
you use? 

Mr. Ingham ; I don't mean just using it for notes; reading books. 
For example, more concrete figures in that line would be 4,000 
people subscribe to the braille book listing. Somewhere between^ 
5,000 and 6,000 read the Reader's Digest, and things like that. 
These are just figures, though, in case you're interested. So 
the figure is increasing. Tnis will be an extremely useful thing 
to have if it will work. 


Chairman Rodgers : Let us now get back to the structuring of the 
project. Dr. Benham, I think, has given us his views. Dr. Foulke, 
did you have anything to add to the structuring of the project? 

Dr. Foulke : No, I don't think I have anything to add. I think I 
could summarize what I would do if I were going to research. 

Chairman Rodgers ; All right, do you want to give us your summary. 

Dr. Foulke : I think I would start out with a legibility study in 
an effort to learn something about the legibility of the dot com- 
binations that are possible with the nine-dot system. 

I think I would attempt, on a gross basis, just to learn 
something about how people would respond to a system such as the 
one that is presently before us; that is, the attempt to teach it 
to some people and get some impression about how well it did appear 
to work, and then I think, probably at the same time, I would 
attempt to construct a brand new code on the most rational basis 
possible, using all of the combinations that are available within 
a nine-dot system on the basis of an analysis of the language, re- 
dundancy, of sequential dependencies, things of this sort and what 
was learned about legibility or ambiguity of the elements in the 
stimulus alphabet. Also I would give some consideration, I guess, 
in the assignment of elements in the stimulus alphabet to elements 
in the response alphabet, 

I would give some attention to the matter of stimulus-resrionse 
compatibility. I think that this would be the kind of research 
program, then, that I would undertake if I were going to do it. 

Chairman Rodgers : Thank you. Mr, Ingham. 

Mr. Ingham : As far as the projects go, I think what has been said 
just about summarizes what I would like to say, except that I would 
reinforce the comment about a gross survey, about, for example, 
having the present code used to do maybe fifty pages or a hundred 
and distribute it to a reasonably large number of people with 
questionnaires, carefully designed to elicit comments on both chara- 
cters and context and understanding. I would do this as soon as 
possible , 

Chairman Rodgers : Mr. Liechty, would you care to summarize your 

Mr. Liechty : The notes that I've taken, I think I've already touched 
on them. I think I will pass up and save the time. Thank you. 


Chairman Rodgers ; Dr. Nemeth, what would you suggest with respect 
to research procedure? 

Dr. Nemeth ; I would suggest that we should undertake several 
directions of investigation: One, produce just ordinary English 
braille with no additional contractions but just the contractions 
of English braille in the nine-dot cell form, and allow people to 
accustom themselves to reading just that. You realize that they 
will thereby only need initial and final-letter contractions in 
this form of the research. They will not have to learn anything 
new essentially but just to acclimate themselves to the change of 
the spacing, and to find out what kind of acceptance or what kind of 
troubles you get there. 

Two, I would make a statistical analysis of the frequency of 
occurrence of words in order to determine what kind of codes 
would be most suitable for them. I think that a one-to-one assign- 
ment of 512 nine-dot sub-sets to the 512 most frequently occurring 
letters or groups of letters would just be fatal. I don't think 
you would get anywhere, 

Mr. Strom ; I agree. It has to be different. 

Dr. Nemeth ; So this is the second thing. Then I would suggest 
that the thing to do is to bring the code that you eventually devise 
gradually to bear, produce material maybe a volume long, and on 
the first two pages you do nothing but Grade Two, and then you 
add three contractions of the code you devise on the next two pap:es, 
and then you add three more contractions, and so on, because learn- 
ing the code and reading the thing all at once is going to get you 
an unrealistic response, 

Mr. Covici ; What if some of the contractions are different? I 
mean you want to keep all the Grade Two contractions in everything 
we did? 

Dr. Nemeth ; No, For example, the contraction for "in"--Bob made 
it three dots wide to avoid the other troubles. The contraction 
for "can", he also made three dots wide for the same purpose. 

Mr. Strom ; So that it could be a part-word sign, that you could 
use it in the middle of a word without its looking like a letter. 


Dr. Nemeth ; I see. I thought it . was for orientation. In other 
words, vou would not wonder whether it was "a" followed by dot 1, 
or whether it was "a" followed by dot k, 

Mr. Strom ; It serves both purposes. 

Dr. Nemeth : Anyway, there are relatively few changes. The "it" is 
another change which can be used as a part-word sign. In other 
words, you could introduce all of Grade Two plus those contractions 
of Grade Two, which are changed and then add brand new ones as you 
go. I'm afraid that while it will be interesting to make the kind 
of experiment that Dr. Nolan suggested; namely, the ease of perceiv- 
ing meaningless dot configurations, I don't think it does begin to 
tell the story of readability. 

Mr. Covici : We do not know. 

Dr. Nemeth : That's all I have to say. 

Chairman Rodgers : Thank you. 

Dr. Nolan : I can only say I disagree! 

Dr. Nemeth : Okay. 

Dr. Nolan : I have already given my comments on the wav I think 
the research should proceed. 

Chairman Rodgers : Yes. 

Dr. Nolan : I don't think I have anything more to add. - 

Chairman Rodgers : You don't have anything else? 

Dr . Nolan : Only to say that I disagree with my friend, Dr. Nemeth. 

Chairman Rodgers ; I was just going to interject a short comment 
which may be a redundancy. As I see it, readability is to a large 
extent a highly ];efined form of discriminability , so that when we're 
studying degrees of discriminability for the purpose of relating it 
to legibility, we have to bear that in mind very carefully. 

Dr. Nolan : That's right. I would go along with that, but legibility 
is a basic ingredient. 

Chairman Rodgers ; That's it, 

Mr. Strom; It's one factor among many. 


Chairman Rodgers : Dr. Scott? 

Dr, Scott : I want to make two brief comments, but I want to 
preface them by saying that I know nothing about braille and so my 
comments are to be taken with that in view. 

One is that I have the feeling we don't really have a great 
deal of research on the six-dot braille cell and this is one of the 
difficulties we have had here, because as questions come up about 
"Nine-Dot Braille" it becomes apparent that we don't even know the 
answers about six-dot braille, and I think it would be a mistake 
if the research strategy were to study only "Nine-Dot Braille", 
because at the end of it we would only know whether or not it 
was a feasible system and if it weren't a feasible system, we would 
have gone through a lot of trouble to determine this. It seems to 
me a much more realistic thing to clue in the basic problems about 
which we have no data at the present time which are applicable 
not only to the six-dot but to the "Nine-Dot Braille", as well. 
In other words, we ought to be studying both systems or at least 
our research ought to be phrased so that it's applicable to both 

The second thing is that there is a analogous experience 
which the field of education has recently gone through and I imagine 
that we might learn something from this experience, because there 
has been an attempt to introduce a new system of reading involving 
various contractions and this is being experimented with by the 
public school systems* 

I don't think we will learn anything in terms of how to set 
up a new system of raised-dot reading, but we might learn something 
about how you go about instituting it or about what kind of -problems 
you meet in doing so, or, if it were decided to use "Nine-Dot Braille", 
I'm sure that the experience of these people trying to institute a 
new system of reading would be very valuable, and also how they went 
about researching it, which would be a very valuable thing in terms 
of setting up strategies here. 

Chairman Rodgers : With respect to that research, of course, I 
think it could be very helpful to us in obtaining all the reports 
on the methods and results thereof. So we're very happy to have 
that comment. 

Now Mr. Strom. 
*(See NEA Journal , "Value of ITA", September 196^ - Pages 20 to 22). 


Mr. Strom ; I go along with the proposals of all the people who 
have mentioned most of the things to be considered already. The 
very fact that we have disagreements in matters of details with 
regard to research strategy, implies that we somehow have to set up 
a structure of a small group of research assistants who are going 
to plan this, and it is somehow going to finally come to a con- 
clusion regarding different ways of going about the problems and 
actually putting them into execution. 

The group has to be small but diversified enough so that it 
contains representatives from the fields of psychology, education, 
braille, mathematics, and whatever other specific categories are 
necessary, and who are fsuniliar with matters regarding machinery 
and with natters regarding testing methods and educational methods. 
Further, this group has to be run sort of semi-independently and, 
therefore, we somehow have to recommend that a small sub-committee, 
say the braille legibility committee, the braille legibility testing 
or something like that, be set up as a result of this conference 
to ascertain what kind of equipment and research is necessary and 
to plan this program and then perhaps solicit funds from the 
American Foundation for the Blind for carrying this out. 

Some small group would need to give basic direction for the 
research efforts, and I think now is the time to decide how we're 
going to go about getting such a group. 

Mr. Covici : The first thing I would do in a research project is 
to review all the literature that's around and find out how valuable 
it is. I don't think anybody has mentioned this but I think it 
should be done, even though I don't know how much there is. 

Secondly, I would set up an autonomous research group. In 
other words, I agree substantially with Robert's statement, but I 
must emphasize that it should not be too closely linked, it should 
not have too many administrators on it. It should be research 
people in different fields who could do a little work independently 
of the group as well as with the group, and so forth. I would set 
up studies; I would develop a code; actually develop the material, 
have people read it, and I would also do some legibility studies 
on the meaning of the combinations just so that some interest 
might come out of it. I would do some information theory, and so 
forth, to develop a code and in general to do what he said, but 


I really must emphasize that we've got to get something out and 
have people read it. What good is it if they don't? 

Chairman Rodgers : If no one else has any comment — and the time 
is gettiniT a little short — I should like to get to the final phase 
of the Conference; namely, the organizational aspect of the structure. 
I thoroughly agree that it would be most unfortunate if no action 
were to result from this second Conference, because this is the 
very reason we called a second Conference on "Nine-Dot Braille". 
I, therefore, am open to suggestions as to what steps should be taken 
to organize the nucleus. I personally feel that this is a highly 
competent group, and think you all could make valuable contributions, 
if you were to organize yourselves into a research nucleus from 
which the whole project would stem and develop. By "whole project" 
I mean the research contents which are now on record. 

You may disagree with that proposal and we will mnke the 
rounds again in the regular alphabetical order. Dr. Ashcroft, 
what do you think about this nucleus idea and about this group of 
participants constituting it? 

D r. Ashcroft : I think that this group might constitute an advisory 
committee. They might be very helpful, but I think ultimately 
the work will have to be done by individuals, and I would like to 
see this group or some group that is constituted for this purpose 
of being an advisory committee publicize the work widely and 
attempt to get the individuals interested who need to be. And then 
facilitate the writing of proposals by the individuals for funding 
by private agencies like the Foundation, although I'm sure that there 
are also federal funds that could be solicited through grant t)ro- 
posals to carry on the work. But it seems to me that ultimately 
the work has to reside in the hands of individuals to get it 

Chairman Rodgers : All right. Dr. Benham. 

Dr. Benham : Whom did you put down on that nucleus? 

Chairman Rodgers : All the participants who are research specialists. 

Dr. Benham : Then I have no other comment. i 

Chairman Rodgers : Dr. Foulke, do you have anything to suggest or 
add to or modify? 

Dr. Foulke ; If we are to set up an advisory committee, I'm sure 
we'll all be happy to advise anybody we could get to take it. I 
don't know if this is the way in which we are to serve as a con- 
sequence of this meeting today, I'm really uncertain about what our 
next step would be. If we were to serve as the group who is actually 
to divide up the problem and define missions and get some research 
on the way, there is going to have to be some money to do it, 


Mr. Ingham ; I think you will have something if that's what 
you're getting at, if you're telling us to go look for somebody » 
I don't know about that. Let me ask a question. There seems to be 
little doubt in my mind that Boh, for example, could polish his 
code along the line that we suggested. The question is suppose 
he did have a code? It's also pretty apparent that we could find 
a way of getting at least a sample prepared of varying material. 

Okay, Now, can anyone in this committee take on the job, 
with a certain aunount of financial support, say, from the Foundation, 
to set up a reasonable research questionnaire, or whatever have you, 
thermoform a very long set of copies, and distribute many hundreds 
of them, carrying on first this gross kind of test, and maybe per- 
haps somebody else in the committee here will be willing to go on 
at the same time carrying on fine legibility research? These two 
approaches immediately offhand could be quickly carried out, I 
think, and at a minimum of cost, I just wonder whether we could 
start on that now. Do we need these as definite avenues to work 

Chairman Rodgers : To whom did you address that? 

Mr. Ingham ; I'm addressing it to the Foundation or the powers 
that be or anyone who would like to take the ball as far as trying 
this research out. 

Chairman Rodgers ; As far as the Foundation is concerned — and Mr, 
Roberts is not here now, so I have to speak on behalf of the 
Foundation — again I have to fall back on what Mr, Barnett said 
this morning. It's on the record. If that is the consensus, if 
that is the recommendation of this group, the Foundation will try 
to do all it can to help. 

I seem to sense that if the problem of funding is solved 
either through the Foundation or through the United States Office of 
Education — 

Mr . Ingham ; Or anybody. 

Chairman Rodgers ; Or anybody else, that this group is ready to 
take on specific assignments with respect to carrying on the 
research projects. 

Mr. Ingham ; Why don't we go around and ask everybody that question? 
I think that would help. 

Chairman Rodgers ; I certainly propose to do just that. 


Mr« Strom : Everybody isn't here now, 

Mr. Liechty ; Dr. Nolan had to leave, Carl, and he sent his regrets 
to you. You were talking and he didn't want to interrupt and he 
said goodbye. He has to catch a plane. Dr. Ashcroft stepped out 
into the library and will be back in a few minutes.* 

Chairman Rodgers ; All right. Then we can proceed in the meantime 
with the others, and the first one would be Dr. Benham. 

Dr. Benham : I don't feel qualified to do more than heckle in any 
research program that might be set up, because there are a lot of 
the aspects of it, psychological and so forth, that are not down 
my line. I could offer assistance in the way of perhaps getting 
some hardware taken care of, or distribution or names of people 
that you might want to send the material to and things of that sort. 
But as far as taking any major responsibility in heading up or in 
working out the research program, I don't think I'm qualified. So 
I will offer my services wherein they are useful, but it wouldn't 
be in the top echelon part of it. 

Chairman Rodgers : By "hardware" you mean reproduction of the 
material; is that it? 

*Dr. Nolan stated in his subsequent letter of June 22nd, 19^4, as 

"Please accept my apologies for having to leave last Friday's 
conference before it was over. I was especially sorry not to be 
able to participate in considering how the research might be under- 

"I have since talked with Emerson Foulke and he described the 
statements of intention to participate made by the members of the 
conference. I certainly feel that APH would be willing to contribute 
by reproducing materials and possibly by making computer time avail- 
able for data analysis. It does not appear to me that individual 
and cooperative efforts of the type agreed upon are the best way of 
energetically studying this problem, A better way would be to set 
up a funded project with its own personnel for just this purpose. 
Project staff could, of course, rely on outsiders for assistance of 
various kinds. This could be done at AFB or through a nearby 
university and would provide continuity that is rarely possible 
through cooperative efforts." 


Dr. Benham ; Yes, or getting machinery of one kind or another 
that you might want. Perhaps even being able to find a little 
money here and there, but not from the point of view of setting 
up an actual research program. 

Chairman Rodgers : The design for the research project itself. 
That in itself is essential to carrying on the work. 

All right, that's on the record. Dr. Foulke, would you 
be agreeable to accept an assignment granted that the funding 
problem was solved, and so on? 

Dr. Foulke ; Yes. If funding were available, I would be willing 
to participate in any of this work, and if I had my choice I 
would prefer to work on an effort to develop an ideal code based 
upon what information is now available and could be collected 
within a reasonable amount of time. 

Chairman Rodgers ; Dr. Ashcroft, the question on the floor now 
is being addressed to each individual participant, and the question 
is: Assuming that the funding problem in carrying out a well 
planned long range research project were to be solved, either 
through the American Foundation for the Blind or the United States 
Government or a combination of private and public dollars, would 
you be willing to serve insofar as contributing to the project? 

Dr. Ashcroft : Yes, perhaps primarily indirectly through scholars 
in our own program at Peabody. A direct contribution on my own 
part would be secondary to the commitments I have, but I would be 
interested in participating on both ways. 

Chairman Rodgers ; I see. Then you would be in a position to 
participate actively in some way. In what particular phase of 

research would you be interested in p^^rtici pating? 

Dr, Ashcroft ; I think primarily at the applied level of actual 
trials of nine-dot codes, field testing, and perhaps if there is 
a place for it, especially in terms of children and the introduction 
in educational programs. One other thing that I would be interested 
in is a programed instructional approach to moving "Nine-Dot Braille" 
into some kind of use. 


Dr. Nemeth ; What approach? I didn't hear, 

Mr. Strom ; Programmed instruction. 

Dr. Nemeth ; I see. 

Mr. Strom ; Perhaps we could get some people at IBM to help us 

out on this. I know a few people there who are interested in braille, 

and in programs. 

Chairman Podgers : Of course, Mr, Covici and Mr. Strom, they would 
not want to participate at all! My point is that I think we know 
what your roles would be. 

Mr. Covici ; Yes. 

Ch-=iirman Rodgers ; The development of codes and so on, right? 

Mr, Strom ; That's right. Also as far as programming computers 
are concerned and that technical area. I mean things like results 
of psychological experiments, program, production of "Nine-Dot 
Braille" according to certain transcription rules that can be fed 
into any machinery that Dr. Benhara or Mr, Ingham can get for me, I 
can work on. 

Mr. Ingham ; I've already indicated that I am very much interested, 
both directly and indirectly, through whomever we can grab at the 
lab at MIT. I can't commit them, obviously. 

Chairman Rodgers ; Dr. Nemeth, are you a research specialist? 

Dr. Nemeth ; Well, I've done research in the past. I would be 
willing to do this much. I would be willing to heckle along the 
literary lines, but when it comes to the mathematical end of it, 
I would be willing to look over any program or even help in the 
development of any program for the writing of mathematics in 
"Nine-Dot Braille" or any other technical kind of materials, 

Mr, Ingham; There couldn't be a better man to do it I 

Chairman Rodgers ; Okay. Dr. Nolan has gone and Carl T. Rodgers 
is not a research specialist! , 

Mr, Ingham ; Bowing out? 

Chairman Rodgers ; Neither a research specialist nor a financier. 
Dr. Scott, 


Dr, Scott ; In view of the objectives of my own study, I've made 
it a policy not to become actively involved in any research programs 
in the field other than those involved in sociology and I, therefore, 
would not be able to become involved actively in the committee, al- 
though I would be interested in monitering its activities and at 
the end of three years when I finish my present project, I would be 
happy to reconsider it. But at the present time I can't accept 
this because of the objectives of my own study. 

Chairman Rodgers : I can assure you that after three years have 
elapsed there will be plenty of work left to do if you want to 
join then. 

Okay. I think we've covered it all now. I'm very, very 
happy to have the specific ways in which this group of participants 
would be able and willing to contribute, because this will con- 
stitute a tangible set of recommendations from which we can proceed 
to try and solve the problems, or at least to look for the people 
that can solve them. Also, if any of you now or at any time can 
offer any suggestions with respect to funding outside the Foundation, 
if you find any money that is available, we would entertain those 
suggestions too, 

I think it has been very fruitful and I was so glad to hear 
the comment — I forget who made it — that, after all, this is research 
and that there is no reason why we should limit ourselves to any 
set of signs or anything else. No one is going to impose anything 
on anybody. We're just trying to learn, find out and then, when 
the appropriate time comes that we know what we're talking about, 
it will be enough time to include the general field and in particular 
those who are responsible for the development and standardization 
of braille codes, because that's very important, too — we have to 
have a stable system — the development and standardization of braille 
codes, literary and special, 

Mr. Ingham : Before you go away, though, what happens next? We 
have all indicated interest and willingness to participate. When 
do we hear from you next or what or how? 

Chairman Rodgers ; I understand that within two weeks or so we will 

get the transcript of this meeting. From it I shall draw up the 

Proceedings which will be available for distribution. At the same 

time I think we have enough concrete material to seek the interest 


of those who would be able to help us, one of them of course being 
the Foundation's administration and its Board of Trustees, No doubt 
we will get suggestions from the Administration as to other possible 
sources of funding, and so we will keep each other in thorough com- 
munication, step by step, so that when the necessary funds are avail- 
able the various phases of research which you have suggested might 
be set in motion. Hopefully, each individual, in consultation with 
everybody else concerned, will take on the phase of research in 
which he has expressed interest in carrying out. 

Dr. Foulke : I would like to add a bit to ray statement. I said 
I would be interested in developing an "ideal" code, and I would 
consider as a necessary step of doing this, a studv of legibility. 

Mr. Ingham : I suppose I'm the only one who hasn't expressed an 

Chairman Rodgers : Did I pass you by? 

Mr. Ingham : No. I just didn't state what I would be interested 
in. Let me just say that I would be interested in approaching our 
group leaders at MIT on the whole project, and for myself I would 
like to take the development of new codes and testing, especially 
the gross testing and distribution and determination of results, 
as well as any instrumentation problems which might arise, because 
there are lots of facilities available at MIT and elsewhere and 
we should be able to tape them. I would be interested in that. 

Mr. Covici : I have one question. Should we set some sort of date 
by which we have some material, because otherwise things might go 
on for a long time. At least some sort of date. 

Mr. Strom : A time for which we have-- 

Mr. Covici : The material to distribute, to send out to people 
to have them try to read it. 

Mr. Ingham : I think before we Can do that, Bob and whoever else 
is going to work on the optimization of this present code have to 
get together, 

Mr. Covici: True. 


Mr, Ingham ; That should be the first thing. The duplication 
problem worked out and the duplications made, and at the same 
time a list of "x" number of people should be available to whom 
this material could be sent, graciously received and looked over, 
and the questionnaire drawn up. 

Chairman Rodgers ; I think Dr. Benham said he could help out 
on people available for that purpose. 

Mr. Ingham ; But he couldn't be the only one, I suspect, because 
much of what Dr. Benham has with his tape library probably consists 
of people who are not regular braille readers, of course. Many 
of them would be of course, but not all of them. His list should 
be supplemented. 

Mr. Covici ; We could do this immediately, 

Mr. Ingham ; Surely. 




The following is a summary of the participants' lengthy dis- 
cussion on this subject. With one exception, the consensus was 
that publicity should not be avoided altogether, but should defin- 
itely be limited to a concise, general announcement, in highly 
professional journals, of the June 19th, 196^, "Nine-Dot Braille" 
Conference and its recommendations to structure and carry out a 
research project on six and "Nine-Dot Braille", 

Dr. Nemeth's reasons for opposing any kind of publicity at 
this time are as follows: 

1. Potential volunteer transcribers will defer becoming 
braillists until the outcome of the proposed research 
is known. 

2. Publicity at this time will engender much opposition 
from the field of work for the blind. 

3. At the present time there is really nothing that can 
be reported to any individual or group. When prelim- 
inary results are available a summary statement can be 
prepared for distribution to interested people. 


Chairman Rodp^ers : Thank you very, very much gentlemen, for what 
I think are most valuable contributions from each of you. 

We are now adjourned. 


August, 1965 -88- 



JUNE 19, 1964 

Submitted by: 






Friday, June 19, 196^ Helen Keller Room 

at 9:30 A»M. American Foundation for the Blind 

9:30-9:50 A.M. Welcoming and Introductory Remarks 

M, Robert Barnett, Executive Director 
American Foundation for the Blind 

9:50-10:15 Resume of Efforts in Developing 

"Nine-Dot Braille" 
Robert Strom, Protagonist 
Student-Harvard College 

10:15 Ten minute break 

10:25-12:30 Discussion : Readability of Nine-Dot 

Cell characters — Type Scale; Tactually 
Discriminable Space Difference of 
Modifying Dots (See Items 1-2, Summary 
Statement of Working Paper). 

12:30-1:30 LUNCH-at the Foundation 

1:30-2:30 P.M. Discussion : Character Configuration; 

Extension and Position in the Nine- 
Dot Cell (See Items 3-6, Summary 

2:30-3:00 Discussion : Space-Saving Efforts 

(See Item 7» Summary Statement) 

3:00 Ten minute break 

3:10-^:30 Concensus of the Participants on 

the structuring of an effective 
research project to determine whether 
or not nine-dot cell characters are 
feasible from the standpoint of: 

a) readability 

b) use for writing by blind individuals 

c) small-and large-quantity duplication. 

Carl T. Rodgers 
Conference Co-ordinator 



At the July I96I Convention of the "American Association of 
Workers for the Blind", in a paper entitled "Some Developments in 
the Various Braille Codes— Literary , Mathematics, Music", I happened 
to raise this question: "In light of the continuing birth of new 
sciences, new scientific discoveries, new methodologies, and, there- 
fore, new terminologies and new forms of established older ones, may 
not the time come when the braille system will begin to burst at 
its orthographic and legibility seams?" 

By taking the number of single dots and dot arrangements (64 — 
two to the sixth power) which the six-dot cell permits as the con- 
stant factor, and the increasing need to represent new terminologies 
and new orthographic styles as the variable factor, we may then con- 
ceive of the whole braille problem as one of trying to crack the 
six-dot barrier. I believe that Mr. Robert Strom's "Nine-Dot Braille" 
proposals represent the first major current effort toward this end. 
Only the most extensive and carefully-structured research efforts 
will determine whether or not a nine-dot cell is tactually and tech- 
nically feasible. 

Preliminary analysis of Mr. Strom's proposals seems to show a 
very close similarity between "Nine-Dot Braille" problems and those 
of regular braille. In essense these are problems relating to read- 
ability of embossed characters as a function of: 

A, Number of dots in a character. 

B, Vertical and horizontal extension of characters, 

C. Character shape. 

D, Characters which are the same in shape, but which 
derive their meaning from rotational position changes 
in the cell. 

E. Characters which are the same in shape but which 
derive their meaning from their upper or lower 
position in the cell. 

F. Type scale (spacing values) 

Q. Space-saving efforts through the use of contractions 
and short-form words. 

A. Number of dots in a character . 

The blind participants who evaluated Mr. Strom's original 
illustrative material at the first "Nine-Dot Braille" Conference said 
that "jumbled" or over-dotted characters cause confusion in discrim- 
ination and recognition. While Burklen ( Touch Reading for the Blind ) 
states that optimal tangibility does not depend on the number of dots 
in a character as much as on such characteristics as simplicity of 


geometric form, and also on what he called "open characters" 
(like the braille letters "m" , "u", or "x"), the tests conducted 
by the Uniform Type Committee of the AAWB indicated a higher degree 
of legibility with respect to characters containing fewer dots, 
from the standpoint of both speed and accuracy ( Proceedings of 
the AAWB Conventions— 1907. 1909, 1911« 1913, 1913 ) » The Committee's 
tests were conducted with American braille and New York Point, the 
two punctographic systems then in use in the United States. The 
relationship of character configuration and number of dots in a 
character to readability awaits scientific investigation on a 
much more thorough level than has hitherto taken place, 

B, Vertical and horizontal extension of characters . 

In its 1907 report to the AAWB Convention of that summer, 
the Uniform Type Committee described a test designed to find out 
whether characters two points high were easier to read than chara- 
cters three points high. The following brief except from the 
Committee's report will enable the group at the Conference to judge 
the value of the test: "In fifty-five trials, the list which in- 
cluded the tall letters was read in one per cent less time, with 
two per cent less errors, than the one in which only the short 
letters were used". Burklen's mention of "the easier perception 
of open characters" (characters which are always three points high) 
may add some support to the Committee's finding on the higher 
degree of readability of the "tall" letters. 

Because a "Nine-Dot Braille" cell would constitute a hori- 
zontal extension of the Louis Braille cell, interest may be added 
to the "Nine-Dot Braille" discussion by citing an excerpt from the 
1961 Proceedings of the Conference on Research Needs in Braille , 
relating to the Uniform Type Committee's investigation of vertical 
versus horizontal signs: 

"At the 1909 AAWB Convention the Committee reported on an 
experiment designed to find out which signs are more legible — the 
horizontal or the vertical ones. Two lists of two hundred signs 
each were prepared. In one of the lists all the signs, varying 
from one to three dots long, were placed horizontally, while in the 
other list the same signs were placed vertically. Each of the 
twelve readers who took part in the experiment, all of whom knew 
both New York Point and American Braille, read the two lists aloud, 


while a seeing person followed the reading with an ink copy, 
marking the errors and checking the time spent on each list. The 
signs were called not by their letter values buy by the number of 
dots they contained. The signs used represented more than one- 
third of the recurrence of the alphabet in New York Point and 
American Braille, The longest of them could be taken as the ex- 
tremes of the two positions — the most horizontal and the most 

"The horizontal signs were read in 33 P«r cent more time, 
with 321 per cent more errors than the vertical signs. Subsequent 
tests gave similar results". 

C. Character shape . 

See Paragraph A. 

D. Characters which are the same in shape but which derive their 
meaning from rotational position changes in the cell . 

Burklen indicated that certain symmetrical characters of 
braille (some 26) differ only in angular position from one another, 
causing reversal reading errors, as in the case of the characters 
composed of dots 1-6, 1-3, and ^+-3. Examination of Mr. Strom's 
revised illustrative "Nine-Dot Braille" material reveals that these 
"mirror characters" may occur with equal or greater frequency in 
the nine-dot codes than in regular braille, as in the case of the sign 
for the word "here", dots 8-1-2-5, which is the regular Braille "r" 
rotated 90 degrees counterclockwise. 

(The nine dots of the cell are numbered: 7*8, 9i downward on the left; 
1,2,3, downward in the middle; and 4,5,6, downward on the right). 

E. Characters which are the same in shape but which derive their 
meaning from their upper or lower position in the cell . 

The Uniform Type Committee designated these as "equivocal 
characters". It summarized its findings as follows: 

"Experiments tend to show that time is consumed and certainty 
diminished by the labor of determining the position of characters 
which are like other characters except for their level in a line." 


Ashcroft ("Errors in Oral Reading of Braille at the Elementary 
Grade Levels") found lower-cell contractions to be fourth (within 
a total rank of seven categories) in the order of difficulty from 
easy to hard, and stated that "lower-cell contractions were the 
most difficult of the group of single-cell contractions." 

There seems to be nothing in the literature in contradiction 
to these findings. This is worth pointing out in view of the fact 
that lower-cell characters occur in "Nine-Dot Braille" at least with 
equal frequency as in regular braille, 

F. Type scale or spacing values . 

This is perhaps the most difficult problem of "Nine-Dot Braille" 
For the original material (see paragraph A) the spacing values were: 

vertical and lateral distance between two dots 

in a single cell O.O8O" 

total lateral-cell space O.I6O" 

distance between corresponding dots of two 

adjoining cells 0.288" 

unknown value between adjoining lines of writing. 

Some of the participants at the first Conference thought that 
the impression of "jumbled characters" resulted from the over-reduced 
spacing values between dots in a single cell. The judgement of the 
participants may or may not have been correct, but it is interesting 
to note that the total lateral cell spacing used for the material 
was O.16O", or O.O7O" greater than the lateral cell spacing (O.O9O") 
of a regular braille cell, and the distance between corresponding 
dots of two adjoining cells was O.O38" greater than that of two 
adjoining regular braille cells, which is 0.250". 

Each of the three "Nine-Dot Braille" slates which were con- 
structed for embossing new material represents a different spacing 
value as follows: 

First slate. Distance between dots in a single cell. . . .O.O85" 
Second slate. Distance between dots in a single cell. . . .O.O9O" 
Third slate. i^istance between dots in a single cell. .. .0.095" 

The spacing between cells and between lines was the same for 
the three slates; namely, O.I6O" between cells and point 0.220" 
between lines. 


As indicated in the "Addendum Sheet", the addition of the 
modifying dots 7, 8, 9» at the left of the cell is unfeasible 
without an increase of "x" decimal of an inch between the modifying 
dots and dots 1,2,3, the mid-cell dots. For instance, the "Nine- 
Dot Braille" contraction for "ast", dots 3t^-7, can be read as 
the regular braille "ch" sign (dots 1-6) followed by dot 4; the 
contraction for ""full", dots 1-2-4-7-8-9, can be read as the 
regular braille "q" followed by dot 4; the contraction for "ound", 
dots l-A--5-7-9» (the regular Grade Two contraction written as a 
one-cell contraction), can be read as the regular braille "m" 
followed by dots 4, '5, and so on. 

The reason for these ambiguities is obvious* The distance 
between dots 3-7 of the "ast" contraction equals the distanice 
between dots 1-6 — the regular braille "ch" sign; the distance 
between dots 1-2-7-8-9 of the "full" contraction equals the dis- 
tance between dots 1-2-5-^-5 — the regular braille "q"; the dis- 
tance between dots 1-7-9 of the "ound" contraction equals the dis- 
tance between dots 1-3-4 — the regular braille "m". 

Many more examples of ambiguities and distortion of regular 
braille configurations can be cited, but it is' believed that enough 
has been said to show the need for an increase of "x" decimal of 
an inch between the modifying dots and the remainder of the cell. 
Since this appears to be true, and since the spacing values which 
were used in the original "Nine-Dot Braille" material seem to 
have disturbed a number of readers at the first Conference, the 
following figures on spacing variables are offered in the hope of 
stimulating more thinking on the problem. 

Regular Braille Cell 

Distance between corresponding dots of two 

adjoining cells 0.250"; 

between corresponding dots of adjoining lines,. 0.400". 

Number of cells in a 10" long line 40. 

Number of lines on 10" high writing space 25. 

Total number of cells per page,...,,,... 1,000, 

Nine-Dot Cell 

(Until otherwise indicated, the distance between two lines 
of writing should be taken to be 0.220".) 


Distance between corresponding dots of two adjoining cells 
when the distance between dots in a single cell is O.O8O", 
and the distance between cells is 0.123": O.283". 

Number of cells in a 10" long line 35. 

Number of lines in 10" high writing space 25. 

Total number of cells per page 875. 

Distance between corresponding dots of two adjoining cells 
when the distance between dots in a single cell is 0.090", 
and the distance between cells is 0.123": 0.303". 

Number of cells in a 10" long line 33. 

Number of lines in 10" high writing space 25. 

Total number of cells per page 825. 

Distance between corresponding dots of two adjoining cells 
when the distance between dots in a single cell is O.O8O", 
and the distance between two cells is 0,l60": 0.320". 

Number of cells in a 10" long line 31. 

Number of lines in 10" high writing space 25. 

Total number of cells per page 775. 

Distance between corresponding dots of two adjoining cells 
when the distance between dots in a single cell is O.O9O", 
and the distance between two cells is O.I6O": 0.340". 

Number of cells in a 10" long line. 29. 

Number of lines in 10" high writing space 25. 

Total number of cells per page 725. 

When the preceding four lateral variables are applied in the 
same order, but the vertical distance between dots is reduced to 
0,072", and the distance between the lines is reduced from 0.220" 
to 0.163", then the number of lines on a 10" high writing space in- 
creases from 25 to 32, and the cell totals per page are as follows: 
1,120; 1,056; 992; 96O. 


The figure 1,000, which is the total number of cells per 
page for regular braille using a 10 x 10" writing surface, might 
serve as the constant factor when studying the readability of 
"Nine-Dot Braille" as a function of "x" spacing variables — lateral 
and vertical. In offering these figures the dot-base diameter has 
been assumed to be 0.053". Changes in dot diameter, and possibly 
also in dot height, might help increase the legibility of nine- 
dot cell characters. 

The foregoing figures were suggestea by the 1920 Report 
of the AAIB-AAWB Commission on Uniform Type for the Blind, and by 
the study on spacing variables of Meyers, Ethington and Ashcroft 
("Readability of Braille as a Function of Three Spacing Variables"). 
Attempts to ascertain the spacing values for embossing New York 
Point — which in some respects resembles the nine-dot cell — have 
thus far been futile. 

Two more considerations with respect to spacing problems 
seem appropriate. The first relates to diagonal dot distances. 
Since both the six-and nine-dot cells are quadrangular, these 
distances may be calculated through the formula "a" square plus 
"b" square equals "c" square. The diagonal distances yielded are 
as follows: 

Regular-Braille Cell 

Distance between dots 1,6 or dots 3»^ 0,201". 

Nine-Dot Cell 

Distance between dots 7, 6 or dots '+, 9 when the distance 
between both vertical and horizontal dots is O.O9O" . . . .0.25^" 

Distance between dots 7, 6 or dots 4-, 9, if the lateral 

distance between dots is reduced to O.O8O" and the 

vertical distance is reduced to 0.072" 0.215" 

The second consideration is concerned with the relationship 
of dots 7, 8, 9, of the nine-dot cell to dots ^, 5, 6 as modifying 
dots in regular braille two-cell signs. In regular braille, words 
and letter-groups are contracted into a two-cell sign by prefixing 
one or several dots to a letter or a contraction — dots 5, 1-5 
for "ever", dots 4-5-6,2-3-4-6 for "their" and so on. 


Query: Aren't dots ^, 5i 6 in effect functionally identical 
with dots 7, S, 9i in a "nine-dot cell", except for a larger space 
between them and the remainder of the signs of which they are a 

If investigation reveals that "Nine-Dot Braille" type, in 
order to be tactually feasible, must exceed the 0.1 square-inch 
surface scale of regular braille, is the solution to the problem 
to be found in a more extensive use of contractions? 

G, Space-saving efforts through the use of contractions and short - 
form words . 

This has been one of the most highly controversial aspects 
of touch reading. Limited investigations by the Uniform Type 
Committee, the Commission on Uniform Type for the Blind, and Ash- 
croft, reveal that certain contraction categories tend to facilitate 
reading, while others tend to produce various types of reading 
errors. There is no reason to believe that in this respect, except 
perhaps for the possibility of substituting contractions for short- 
form words, "Nine-Dot Braille" contraction problems would be any 
different from those of regular braille. Only through additional 
and extensive scientific investigation can one hope to arrive at 
suitable solutions to the whole touch-reading problems. One would 
also hope that in the continuing development of embossed systems 
for blind persons, this principle will always be rigidly follows: 

It is the needs of the blind reader — not of the system that must 

be met . 

The enormous scope of the subject in itself has precluded the 
possibility of a truly suitable working paper. For one thing, only 
a cursory review of the investigation records available was feasible, 
since any well-structured research project would require a far 
lengthier review of the literature than could be presented here. 
The purpose of the foregoing material has been simply to attempt 
to set up the kind of questioning frame of mind that it is hoped 
fully will lead to a fruitful, analytical discussion at the on- 
coming "Nine-Dot Braille" Conference. 




1. Readability of nine-dot cell characters as a function of type 
scale — dot size, dot height, lateral and vertical distances 
between dots in a cell, distance between adjoining cells, dis- 
tance between adjoining lines of writing, 

2. Readability of nine-dot cell characters as a function of 
tactually discriminable space difference between dots ?♦ 8, 9i 
used as modifying dots at the extreme left or right side of 
the cell, and the remainder of the cell. 

3. Readability of nine-dot-cell characters as a function of 
character configuration and character extension--lateral and 
vertical . 

4. Readability of nine-dot-cell characters as a function of 
characters which are the same in shape, but which derive their 
meaning from their lateral position in the cell — left, middle, 
right -nosition; left-middle, middle-right position, 

5. Readability of nine-dot-cell characters as a function of chara- 
cters which are the same in shape but which derive their mean- 
ing from their upper or lower position in the cell, 

6. Readability of nine-dot-cell characters as a function of 
characters which are the same in shape but which derive their 
meaning by being rotated as much as 3^0 degrees. 

7. Readability of nine-dot-cell characters as a function of greater 
space-saving efforts than those prescribed in the "AAIB-AAWB 
Braille Authority's English Braille-American Edition 1939- 
Revised 1962". 

« * 1^ iK « 4i 



JULY 6, 1962 

Submitted by: 

Carl T. Rodgers 
Program Specialist in Braille and Other Tactual Aids 


Miss Kathern F. Gruber, Chairman 
Director of Division of Program De- 
American Foundation for the Blind 

Mr, Kenneth Ingham 
Mathematician-Graduate Student at 
Brandeis University 
Waltham, Massachusetts 

Mr, Howard M, Liechty, Editor of 
the "New Outlook for the Blind" and 
Member of the Board of Trustees of 
the American Foundation for the Blind 

Mr, Carl T. Rodgers 

Program Specialist in Braille and 

Other Tactual Aids 
American Foundation for the Blind 

Mr, George F. Meyer, Executive Director Dr, James Slagle 

New Jersey State Commission for the Blind Research Associate 

and Member of the Board of Trustees of Massachusetts Institute of 

the American Foundation for the Blind Technology 

Dr, Thomas A. Benham 
Professor of Physics 
Haverford College 
Haver ford, Pennsylvania 

Mr, Robert Strom 

Inventor and Protagonist of the 
"Nine-Dot Braille" and an 
undergraduate student at 
Harvard University 

Mr. Robert A, Bowers 

Leadership Fellow 

American Foundation for the Blind 

Mr, Arthur L. Voorhees 

Program Specialist in Vocational 

and Rehabilitation Services 
American Foundation for the Blind 

Mr. John K, Dupress, Director of 

Technological Research 
American Foundation for the Blind and 
Research Associate of the Department 

of Electrical Engineering 
Massachusetts Institute of Technology 

Dr. Everett E, Wilcox 

Program Specialist in Education 

American Foundation for the Blind 

Miss Joan M, Stapleton, Secretary 
American Foundation for the Blind 


Mr, Ted Glazer 
Manager, Systems Division 
Burroughs Research Center 
Paoli, Pennsylvania 

Mr. M, Robert Barnett 

Executive Director 

American Foundation for the Blind 

Mr, Louis H. Rives, Chief 

Division of Services for the Blind 

Office of Vocational Rehabilitation 

Dept. of Health, Education, and 


Washington 25, D.C. 



This Gonferenc* was held on July 6, 1962, from 10:00 a.m. to 
^:00 p.m. in the Helen Keller Room of the American Foundation for 
the Blind. Miss Gruber opened the Conference by stating its 
purpose as follows: 

One of the major responsibilities of the American Foundation 
for the Blind is to evaluate all ideas, systems, methods, 
aids, etc., that are formally ^resented to the Foundation 
for this purrDose. Within the past two months, Mr. .Robert 
Strom -presented to the Foundation a complete system of 
"Nine-Dot Braille" for evaluation. The Program Sn^^cialirt 
in Braille, Mr. Carl T. Rodgers, made the pr^^liminary in- 
vestigations, and as a result of his findings, he recommended 
to the Foundation administration that a small group of tota''lv 
blind re=5ders be convened for the pur-nose of further analysis 
and ex-nloration of this "Nine-Dot Braille. 

Miss Gruber then established a most satisfactorv auditory 
rap-port araonfr those present by having each of them make a brief 
statement of self-identification. Following this, the Program 
Specialist in Braille described the content and the arrangement 
of the Conference material; told how it had been prepared and d\- 
plicated: explained some of the technical difficulties which were 
encountered in producing the material on a somewhat crude prototype 
"Nine-Dot Braille" slate; and stressed that all of the samples and 
the spacing values used were tentative and subject to whatever 
chanp-es and revisions mip-ht be indicated by reader reaction and 
adequate research studies. 


1 . Legibility and Spacing Values of "Nine-Dot" Characters 

It was observed by the participants that "bunching of dots" and 
widely-spaced, isolated dots decrease reading facility. Among such 
signs were the "anv" sign (dots 1-2-3-^-6-7-8-9), the "ity" sign 
(dots 1-3-^-^-6-7-8-9), and the "an" sign (dots 1-9). 

It was also noted that "Nine-Dot Braille" characters which did 
not "overcrowd" the sense of touch a>"e more easily read. Among 
such symbols were the "is" sign (dots 2-^-9), the "ate" sign 
(dots i-8-9), and the "ly" sign (dots 1-2-3-7-9). 


It was suggested that some of the signs which appeared "bunched 
together" might be made more legible by appropriate alterations 
of the spacing values of the nine-dot cell, and it was agreed that 
o-otimal spacing values would have to be established through ad- 
equate research. It was stated that the findings of the study en- 
titled "Readability of Braille as a Function of Three Spacing 
Variables" by Meyers, Ethington and Ashcroft was a fine contribution 
to the field of research in braille and a good starting point for 
further investigations concerning suitable spacing values, dot 
diameter, and dot height. 

2. The Writing of "Nine-Dot Braille " 

Mr. Strom's "Nine-Dot Braille" slate was examined by the par- 
ticipants. It was pointed out that dot 5i being the middle point 
of the square, (the nine-dot cell) might be difficult to write, 
to which Mr. Strom, who had punched out all of the "Nine-Dot Braille" 
Conference material replied that he had encountered no difficulty 
with the writing of characters containing dot 5. Brief mention 
was also made of a keyboard suitable for writing "Nine-Dot Braille" 
as well as for manual operational requirements. 

3. " Nine-Dot Braille" Structure 

The Nine-Dot cell was developed simply by adding the column of 
dots 7,8,9 to the right of the conventional braille cell. It was 
suggested at the Conference that the added column might be placed 
at the left, rather than at the right of the conventional cell for 
the purpose of using combinations of the dots of the added column 
much the same as dots ^, 5« 6 of the standard cell are presently 
used in compound symbols. This would retain, more compactly, the 
already existing symbols of Grade 2 braille, avoiding obsolescence 
and affording the possibility for the creation of new symbols. Mr. 
Strom and the Program Specialist in braille stated that this al- 
ternative for develo^ ing nine-dot characters might be quite feasible 
since the nine-dot characters would then be truly an extension of 
the six-dot cell characters. Mention was made by Mr. Meyer of the 
possibility of a variable cell in order that some of the characters 
might occupy less space than the full nine-dot cell. It was felt 
by some of the participants that this idea, which would be feasible 
through a properly constructed slate or machine, might serve to 
save considerable space and lend continuity to the legibility of 
the characters. 

^. Purposes and Needs of Braille 

Mr. Meyer, addressing himself directly to the Program Specialist 
in braille, asked him to state his "impressions" concerning braille 
needs. The following summarizes Mr. Rodgers ' effort to answer Mr. 
Meyer's request. 


"First, I believe that as a system of touch reading and 
writing for blind persons, braille is a multi-purpose system. 
When we speak of 'needs', what needs are we talking about? Are 
we talking about the needs of the reader who wants to read for 
recreation only? Are we talking about the needs of the *John 
Covicis' who need an orthographic system for the accurate re- 
cording of current mathematical and scientific concepts and data? 
Consider the orthographic discrepancies that exist between the 
symbols provided for in Braille Chemical Notations and How to 
Use Them by Loomis and Mitchell, and the symbols of the Nemeth 
Code. At the present time, braille mathematical and scientific 
notation is in a state of complete fuidity. Are we talking about 
the needs of the blind musician, whose present braille notation 
is highly technical and multi-celled? When speaking of 'needs' 
we have to break them down the way we do when dealing with the 
needs of any handicapped group -- into highly individualized 
needs. Each available medium — braille, talking book records, 
tapes — must be used to meet the needs of individuals . " 

The rest of the discussion on the question of needs revealed 
the feeling that a nine-dot system would be welcomed by "the 
blind individual seeking a way in which to record highly special- 
ized work", but that "the, individual who reads non-technical 
braille is going to rebel," 


The following is a summary of the ideas offered by the par- 
ticipants to give order and meaning to the potential creation of 
"Nine-Dot Braille": 

1, Wherever possible there should be strict adherence to 
the principle of orthographic representation (each symbol 
to be assigned one meaning). 

2, Concentrate on developing characters not now available 
in braille for representing certain existing ink-print 
characters . 

3, While absolute standardization of symbols and symbol 
meanings does not really exist in inkprint, it would 
seem wise to establish code uniformity by providing 
braille equivalents for those inkprint symbols used 

most commonly and having the least variations of meaning. 


h. Whenever possible use the right side of the "Nine- 
Dot Cell" for the special symbols of mathematical 
and scientific notation, using the left side of the 
cell only when it becomes necessary to avoid con- 
fusion with a symbol of the literary code. 

5. New characters should be created according to a 
consistent pattern or system of development, 

6. "Nine-Dot Braille" characters should be developed in 
keeping with logical inkprint presentation. For 
example the representation of the inkprint notation 
which means "a-sub-one" should be designed so as to 
read exactly like that, and not "a-one-sub." 

7. In developing "Nine-Dot-Braille" make it as compatible 
as possible with the old system so as to reduce the 
necessity of learning an entirely new system. 

8. The establishment, through scientific experimentation, 
of optimal spacing values as an indispensable re- 
quisite to the legibility of nine-dot characters. 

9. Experimentation with other types of cells, such as 
2 points vertical and k points horizontal, etc., 
should be included in future research on touch reading, 



For the purpose of illustrating the urgency for creating a system 
of touch reading and writing that can adequately meet the require- 
ments of modern symbology, Mr, Strom read a passage from the book 
Principia Mathematica, a reading "must" for students of higher 
mathematics, and which cannot be made available to blind mathemati- 
cians because the existing mathematical braille notation provides 
no orthographic equivalents for the symbols used in this book. Mr. 
Strom stated that suitable single-cell equivalent svmbol":^ could 
be provided by "Nine-Dot Braille", for not only Principia Mathematica 
but mathematical logic as well. " 

The material prepared by the Program Specialist in Braille in- 
cluded the four following questions, the answers to which sum up the 
participants' consensus of opinion on what action, if any, the 
American Foundation for the Blind might take regarding Mr. Strom's 
"Nine-Dot Braille" proposal: 

1. In general, does it appear that the reading finger can s-oan 
and comprehend Nine-Dot characters? 

Dr. Benham Yes, 

Dr. Slagle Yes. 

Mr, Dupress Yes, 

Mr. Ingham Yes, but samples of "Nine-Dot Braille" should 

be produced with more adequate writing equip- 
ment in order to be able to judge better, 

Mr. Meyer Yes, it could be. 

2. Does it appear that some of the characters are especially diffi- 
cult to read? 

Dr. Benham Yes, some are hard to rend, but if dot height 

and spacing values were uniform, they would be 
easier to read. Uniformity and better character 
designing (placing indicators before instead of 
after the basic symbols) would, I think, make 
the system easy to read, 

Dr, Slagle Yes, characters that have many dots are diffi- 
cult to read. The use of dot 5 might, in 
general, be avoided when you have a choice. 
Sparsely-located dots cause difficulty, 

Mr. Ingham Yes, characters with too many dots are hard to 

read. Also, the column of dots 7*8, 9i might 
make for easier readinp if added at tne left 
instead of at the right of the basic cell, 


Mr, Meyer Yes, difficulties increase when you consider 

adjacent characters. Because of the spacing 
values which have been used for this material, 
you can't tel] where one cell leaves off and 
the other begins. 

Mr. Dunress — I will agree with the group that where there 

are only a few dots, left or right, the charac- 
ters are difficult to read. There is need for 
research on this, 

3. Would it be desiral.le to treat the "Nine-Dot Cell" as an ex- 
tension of the "Six-Dot Cell" in developing nine-dot symbols? 

All a.p-reed in the affirmative, since this would avoid render- 
ing the old system obsolete, would eliminate the need for re- 
learning many new characters and would, therefore, result in 
potential subjects being more willing to cooperate in field- 
testing projects, 

^, From the standpoint of present-day orthographic needs on the 

one hand, and of tactual legibility requirements on the other, 
does the "Nine-Dot Braille" concept appear sufficiently pro- 
mising to justify undertaking the extensive and comprehensive 
experimentation, field-testing, and other long-range steps 
necessary for the development of the proposed concept into an 
optimal, multi-purpose system of touch reading and writing? 

All agreed that additional investigation must precede any large- 
scale experimentation and other long-range action. As part of 
further -oreliminary investigation, the participants recommended 
a better-constructed, more sophisticated "Nine-Dot Bra-'"'!*" 
slate to afford better evaluation of the readability of nine- 
dot characters. A few Foundation people and a few well 
selected outsiders should continue the investigation. Further- 
more, a second conference on "Nine-Dot Braille" was recommended, 
at which time a definite decision could be made in answer to 
question ^, 

The protagonist of "Nine-Dot Braille" posed the ouery: "Do 
you think v/e can recommend experimentation as to the size of 
the cell and s];)ecific values?" The unanimous reply was 
"Definitely Yes". 


Finally, everyone recommended gradual and thorough investi- 
gation, evaluation and data gathering of the proposed system, with 
emphasis on technical needs for the benefit of the blind student 
of higher mathematics in particular and technical training in 
general. Stress was laid on making no announcements or taking any 
action which might lead to a renewal of a "War of the Dots" or 
"Battle of the T;/-Des" among the users of braille or among the 
workers in the field. 




Submitted by: Dr. Abraham Nemeth 

Assistant Professor of 



Submitted by: Dr. Abraham Neraeth 

Assistant Professor of Mathematics 


Spacing Parameters 

Betv/een dot centers within a single cell OQO inches 

Between right side of one cell and left side 

of next cell l60 inches 

Between bottom of cell on one line and top of cell 

on next line 220 inches 

In normal six-dot braille, it is possible to emboss 40 cells 
to the line and 25 lines to the p-^.ge if the dimensions of the p ige 
are 11 inches by 11>^ "'nches while still maintaining suitable margins. 
With the same spacing parameters for "Nine-Dot Braille", it is poss- 
ible to emboss 29 cells per line and 25 lines per page while still 
maintaining suitable margins. There thus 1,000 cells per page 
available in six-dot braille with 725 cells per page avail- 
able in "Nine-Dot Braille", Proper comparison of the two systems 
requires that the spacing parameters be assumed equal in both. 

It is assumed that ordinary adult prose has been transcribed 
in the two systems being compared, .Because of the variable length 
of words and syllables, it is not possible to emboss either all of 
the 1,000 cells per page in six-dot braille or all of the 725 cells 
per page of "Nine-Dot Braille", It is estim.^ted that an average of 
1.6 cells per line are unused at the ends of lines for these reasons. 
Accordingly, there will be 40 cells per page lost in either svstera. 
There will thus be 960 cells per page available for six-dot braille 
and 685 cells per "page available for "Nine-Dot Braille", The 40- 
cell-per-page looS constitutes a 4 per cent loss in six-dot braille, 
but a 5.5 per cent loss in "Nine-Dot Braille", We will continue to 
assume, from this point onward, that there are 96O cells and 685 cells 
per page, resDectively , available for braille transcription in the 
two systems. 

It has been pro-nosed that a nine-dot cell be regarded as com- 
posed of a left third to accommodate either com.position signs or 
"prefixes" of two-cell contractions, and a right two-thirds to accom- 
modate letters or part-word signs affected by such composition signs. 


or the "bodies" of two-cell contractions. Four of the six 
composition sip:ns of English Braille can be accommodated in the 
left third of a nine-dot cell. These are the capital, italic, 
letter, and accent signs. The number sign and the termination 
sign cannot be accominodated in this way. 

Each two-cell combination of composition sign followed by 
a single letter or part-word sign, or of a prefix followed by 
the body of six-dot braille can be accommodated in a single cell 
of "Nine-Dot Braille", so that there is a gain in "Nine-Dot 
Braille" of one efficiency unit. In order for "Nine-Dot Braille" 
to be more efficient than six-dot braille with respect to the 
amount of s-nace occupied, "Nine-Dot Braille" must score a gain 
in excess of 275 efficiency units, since this is the number by 
which the cells per page in "Nine-Dot Braille" falls short of 
the cells per page in six-dot braille. 

In English Braille, Grade Two, there are ^7 two-cell 
contractions. A survey of adult prose material suggests that 
the frequency of occurrence of such contractions is at the rate 
of about 25 per page of six-dot braille. Since there are seven 
prefixes and 5^ bodies available, it is theoretically possible to 
construct 7 x 56 or 392 two-cell contractions, as in Grade Three. 
It will be assumed that the rules for forming contractions can be 
formulated in such a way that no conflicts arises from the dual 
role played by four of the prefixes as composition signs. Whether 
the increased number of contractions acts to slow down or to 
speed up the reading;- rate is a psychological matter for investi- 
gation, and is entirely disregarded here. Thus, when 3*^2 is 
compared with ^7i it is seen that the incidence of two-cell 
contractions could be about eight and one-third times as great 
as is the case in grade two. It will therefore be assumed tViat 
the density of two-cell contractions would be eight and one-third 
times as great when using all the possible 392 two-cell contractions 
than is the case when using the 47 two-cell contractions of English 


It may therefore be expected that the frequency of occurrence 
of two-cell contractions would be at the rate of (8 1/3) times 25 
or about 209 occurrences per page of six-dot braille. This is a some- 
what generous assumption, since it is likely that the first ^7 two- 
cell contractions of English Braille have a greater frequency of 
occurrence than the remaining possible 3^5 two-cell contractions. 

A survey of adult rirose material suggests that the occurrence 
of the capital, italic, letter, and accent signs combined is at 
the rate of about 15 per page of six-dot braille, grade two. If 
the fully contracted system of 592 two-cell contractions were used, 
these composition si(2;ns would occur in 751 (that is, 9^0 - 209) 
cells instead of in 9^0 cells, as in grade two braille. At the 
rate of 1^ occurrences in 751 cells, there would be 19 composition 
signs per page of highly-contracted six-dot braille. 

The occurrence of the double capital sign or of the double 
italic sign might also be taken into account. However, the single 
capital sign followed by a two-cell contraction can result in a 
gain of only one, and not two, efficiency units in nine-dot braille. 
Because these two factors tend partially to cancel each other, and 
because the occurrence of either is relatively rare, we neglect both 

The 209 occurrences per page of two-cell contractions in highly 
contracted six-dot braille, plus the occurrence of 19 composition 
signs per page result in a gain of 228 efficiency units when the 
same material is embossed in "Nine-Dot Braille". As was indicated 
earlier, a gain of 275 efficiency units is required just to break 
even. The gain of 228 units of efficiency thus falls short of the 
break-even point by ^7 efficiency units. We must therefore conclude 
that if the same spacing parameters and the same set of two-cell 
contractions are used both in six-dot and in "Nine-Dot Braille", 
six-dot braille still has an advantage over "Nine-Dot Braille" with 
respect to s-nace requirements. 

An alternative suggestion has been made that the "prefix-body" 
concept be abandoned, and that each nine-dot character be made to 
represent a contractable combination in its own right. From this 
point of view no more contractions are available than in the prefix- 
body concept, and its application can only result in a permutation 
of the available nine-dot characters for the representation of the 
desirable contractable combinations. 


If each of the spacing parameters is reduced by .010 inches, 
it would be possible to emboss 42 cells of six-dot braille per 
line compared with 31 cells of "Nine-Dot Braille" per line. It 
would also be possible to emboss 2? lines per page in each system. 
With the new parameters there would thus be 1,13^ cells per page 
of six-dot braille compared with 837 ce]ls per page for "Nine- 
Dot Braille". Again allowing a loss of 1.6 cells at the end of 
each line due to the variable length of words and syllables in 
each system, there would be available 1,091 cells of six-dot 
braille per page compared with 79^ cells of "Nine-Dot Braille" 
per page and the break-even point would come with a gain of 297 units 
of efficiency. The increase from 960 cells to 1,091 cells per page 
represents a gain o-^ about 13.7 per cent. If th^re are 128 occur- 
rences of two-cell contractions and composition signs usinp- the 
standard parameters, then, at the same rate, there would be 25Q 
such occurrences using the smaller spacing parameters. "Nine-Dot 
Braille" would thus provide 25^ units of efficiency per page but 
this is still 38 units of efficiency short of the break-f-ven -^ojnt. 
We must again conclude that six-dot braille still rpnain.-^ no--f^ 
efficient than "Nine-Dot Braille" with resp-ct to s-ace rem i -rprr^nts 
even when the smaller parameters are used. 

The apparent advantage of "Nine-Dot Braille" over six-dot 
braille which is evidenced in the sample materials distributed 
to the conference participants comes from the fact that the nine- 
dot transcriptions were made using both the smaller parameters as 
well as a hij:;hly contracted system, whereas the six-dot braille 
with which it was compared used standard parameters together with 
the standard contractions of Grade Two, 



As is well known, many of the char'.cters of six-dot braille are 
intrinsically ambiguous, and that their precise orientation within 
the cell is often indicated by the placement of a full cell of dots 
in an adjacent position. In the reading of standard prose material, 
such ambiguity is almost entirely resolved both by the presence of 
adjacent characters which serve as orienting symbols, and by the 
context of the material itself. What role each of these factors 
rlays ir the resolution of these ambiguities is not known. We shall 
here be concerned only viith intrinsic ambiguity, and both of the 
factors which resolve it will therefore be absent. 

Characters vary with respect to their degree of ambiguity. Thus 
a one-dot character has six degrees of intrinsic amhie-uity; the 
letter b has four degrees of ambiguity, since, without orientation 
and context, it could be taken as dots 1-2, 2-3, ^-5i or 5-6; the 
letter c has three degrees of ambiguity since, without orientation 
and context, it could be taken as dots 1-^, 2-^, or ^-6; the letter 
e has two degrees of ambiguity since, without orientation and context, 
it could be taken as dots 1-5 or 2-6; the letter m has but one degree 
of ambiguity, since it can only be taken as dots 1-3-^. It is thus 
clear that a rating of one constitutes no ambiguity at all. It is 
proposed, then, to neglect the space, and to rate the ambiguity of 
each of the remaining 65 characters of six-dot braille and the remain- 
ing 511 characters of "Nine-Dot Braille". The average of these ratings, 
obtained by dividing the sura of the ratings nn each svstem by 63 and 
by 511, res-oectively , will then be a measure of the intrinsic ambiguity 
of each system and these intrinsic ambiguities may then be comp^'red. 

We use the letters A, B, and C to designate the dots in the l^-ft 
third of a nine-dot cell from ton to bottom, respectively; the numbers 
from 1 to 6 are used for the remaining dots of the nine-dot cell in 
complete correspondence with the dot numbering of six-dot braille. In 
"Nine-Dot Braille" a one-dot character ha- nine degrees of ambiguity; 
a character consisting of dots A-B has six degrees of ambiguity since 
it could be taken as dots A-B, B-C , 1-2, 2-3, ^-5, or 5-6; etc. 

While it is nossible to assign an ambiguity rating to each 
character of each system by inspection, it is also possible to develop 
a formula for t^-is purnose. With each character there is associated 
a certain number of exterior blank columns and exterior blank rows. 


To find the number of exterior blank columns associated with anv 
character, start from the left edge and from the right edge of the 
cell and count the number of blank columns which are r^resent. For 
example, the character which consists of dot 1 in six-dot braille 
has one blank exterior column; the same character in "Nine-Dot 
Braille" has two blank exterior columns. The character consisting- 
of dots 1-3-^ has no blank exterior columns in six-dot braille, but 
has one blank exterior column in "Nine-Dot Braille". The charac- 
ter consisting of dots A-C-4-6 in "Nine-Dot Braille" has no exterior 
blank columns althoup-b the middle column is blank. In a similar wav, 
to find the number of exti^rior blank rov/s associated with a charac- 
ter, start from the top edge and from the bottom edge of the cell, 
and count the number of hi ark rows which are present. For example, 
the character consisting of dot 1 has two blank exterior rows in 
both systems, the ch-^.racter consisting of dots 1-^-^-6 has no blank 
exterior rows in either system even though the middle row of this' 
character i.s blank. 

We let c denote the number of blank exterior columns associated 
with a character. In six-dot braille, c can have one of the values 
or 1; in "Nine-Dot Braille" it can have one of the values 0, 1, 
or 2. Similarly, we let r denote the number of blank exterior rows 
associated with a character. In both systems r can have one of the 
values 0, 1, or 2. If R denotes the ambiguity rating for any charac- 
ter, then the formula for "R is R=(c +l)(r + 1), For example, the 
character consisting of dot 1 in six-dot braille has associated with 
it one blank column and two blank rov/s. Therefore for this charac- 
ter, R = (1 + 1)(2 + 1)=2 X 3 = 6. The same character in "Nine-Dot 
Braille" has associated with it t/o blank columns and two blank rows, 
so that R = (2 + 1)(2 + 1) = 3 X 3 = 9 for this character in "Nine- 
Dot Braille", The formula for R holds equally well for any rect- 
angular cell such as the two-by-two cell of New York Point and the 
four-by-two cell of the German shorthand system. 

The tables which follow list the 63 characters of six-dot braille 
and the 311 characters of "Nine-Dot Braille", together with the 
ambiguity rating of each character in each system. It is found that 
the average ambiguity for six-dot braile is 2.05 and that the corres- 
ponding rating for "Nine-Dot Braille" is 1.72. We, therefore, conclude 
that "Nine-Dot Braille" has less intrinsic ambiguity than six-dot 
braille and that the ratio of the ambiguities is 1.72/2.05, or .84; 
"Nine-Dot Braille" is only .84- as ambiguous as six-dot braille. 




NOTE: In the following table the column headed (1) specifies 
the six-dot braille characters by dot numbers; hyphens are omitted 
between these numbers. The column headed (2) specifies the ambiguity 
ratins- of the corresron ding character in column (l). 

(1) (2) (1) (2) (1) (2) 





























' 1256 


































































































Total rating 





















B • 


S ^ 


iH H 
O ^-' 






C B 


X! 3 


Eh H 






10 c 
ti -H 






A U 






3 +J 



C O 






4) ^( 



W {C 


C A 


A O 









c c 


«) -H 



«J TJ 



^ E 



+J o 


c p. 



^ CD 






-d tn 



*) t^ 




-P o 




-P o 





t — ' 


E c 



o x: 








U <n 



to o 



to bo 



c c 



*) -H 



A +J 

(— 1 


P, «0 


>> u 








.- 4-) 


ro -H 


^H 3 


4) ho 


XI -H 




3 S 



4J C 


o x: 

T3 -P 



>> 10 

XI « 


• • 

eo <M 


;h -h 


«) o 


+i 4) 


O P4 

nS (0 

X3 nj 









f\J N^i J- lAVO r<^^ lAMD ^lAV£)mvX>VOnjKN^lAv^hrxJ-irN 




VO ITNVOVO (\J KS^ lAVD K\4f lAVD^ ir\»vD lAV£)>sO r^^ lAvO 



O r-l (M r<^ ^ IAV£) iH <M K>^ lAVO f\J rr\ ^ ir\*Xi r<~\-d- lAVD J- 

0^a^0^a^0^a^O^O^a^VO KNVDJ- f\j hf>(\l HVX)^vi5^ OJ K~\ 

„ PQ O H OJ KN-d- ITNVT) O H f\J KN J- IA 

PQ O H OJ K>^ lAVO <a;<C«a;<i;<a;<!«aJ<ajpQpQpQpqpQpQ 



>sD vX) v^ ^£> LTx VT) i^£) 

Lr\ ltn ir\ lA J- J- lA 






LTn VX) VX) \X) -d- ITNMD 

J- ^ lA ir\ K^ r<~\ tox 

•sDd" lAO m^sOvO KNcf LTWDd- lA^D lAVDV£l-:f Lr\V£) LTN^X) r<~\rAK^iJ- OJ OJ fV' 

(\]K-xfAK\-d-j-Lr\nj(\i(\jf\jK\K\N--, j-j- L^^K^K^K^J-^ooooHr^H 








kd [Tskd \d ^ ^ ir\ 
cfiAvoirv^vDj-ir\v£)ir\v£)>sOmv£)vO'X)OOOOuor\jK\N~, NAj-ojoj 











■ - - lAvO irwO VX5 4" LA 

lA PA-d- -4- LA rA N^ 



H r-l H 

v«.«.« ^-^ ^*^ lAVOvX) lAVOVOvD lAVDvOMDVD ITvVOVDVOvDVD lr^VD^^VDvr) 




H H H H H H H 

r-i t-i r-\ r-i 








rH H iH H H 




VOIAVDvD lAVO>^VO irNv£)>X)V^vr> 4- lAvO lAvDVD lAvOv£iVO lAVD VX) VD VO [TwD 
^■^ •3'. "^^ J-lTMA-d-^-irMTMrvN^KN Kn J- 4- IT, J- 4" ITN ITnJ- J- lA IT, lf\ J- J- 



iH H H iH iH H 


(H H H H iH 




K>cr4- ir\4--d- irsir\rr\f<>K\^j- ir\^^ iAir\4-j- irstr\ir\ooooc_5oo 





vi3vX)ir\v£)vDv£)V£iv£)v£)^-d--:l- LT, IT, Lr\Lr\miALrN^ 

LTx LP <f ^ LA IT > iTN L^^ LP- ir\ r^ K~', r^ -d- -cf -d- J- -d- -d- f^ 
J- J- K> K, K- J- -4- -c^ -d- -4- (M fv oo rv r<~x r^ K> K\ K^ fM 


H H iH H iH 






OJ rj fV' Tv: (\J H H H (H H rH f— I H rH H I— I I— I rH H 






VD vo VD vx. f\J 

J- LAVDJ- LP, V£) U^,V£)V£)-:t- LAVX) ITWOVX) 

OJ f\i f\. 
rH rH 

fi- t^ t^ 


LT^ K\ K\ r<^ ^ -:*■ LA 
rH fVi (\J f\J OJ f\J f\J 




IN C- 





























j-cj- ir\iAmr\j cm c\j (\j r^rAi^. -:t -d- lakxknknj- ^ 

K-. K~^K^J-4-lH.H^HrHr^r^HHHHa)(^Jf\JC\JC\J 
(\|(\jrvif\JI^. 0C1OO0O0UU00000O 



H H H 

H r-H rH rH 

OJ ru OJ (H 


LPWT)^ invX) lAVDvO K\^ lPvVX)J- U^VD IAvX)VC-:l- IPWD 



K71673 c.-i 


Proceedings, N'i'C AF:B, June 19, 
196h. ' ' 

Date Due 

MAY !e«?